<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Tony Wu on Medium]]></title>
        <description><![CDATA[Stories by Tony Wu on Medium]]></description>
        <link>https://medium.com/@tonygwu?source=rss-495ebce96386------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*rYMc0Q2j3AxvaTLL.jpeg</url>
            <title>Stories by Tony Wu on Medium</title>
            <link>https://medium.com/@tonygwu?source=rss-495ebce96386------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 20:09:25 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@tonygwu/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Smarter Way to Prep for System Design(Backend)]]></title>
            <link>https://tonygwu.medium.com/smarter-way-to-prep-for-system-design-backend-d89b10991151?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/d89b10991151</guid>
            <category><![CDATA[system-design-interview]]></category>
            <category><![CDATA[system-design-concepts]]></category>
            <category><![CDATA[interview]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Thu, 04 Jan 2024 00:39:27 GMT</pubDate>
            <atom:updated>2024-01-04T00:39:27.131Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Challenge</h3><p>If you’re an interview candidate prepping for a System Design as a Backend Engineer, there’s a lot of potential material to cover. When it comes to preparation for technical interviews, I divide skills into two categories:</p><ol><li><strong>Substance</strong>: Your actual technical proficiency and understanding.</li><li><strong>Interview Technique</strong>: Your “interview skills” or how well you manifest your technical abilities in an interview setting.</li></ol><p><strong>Note</strong>: This post will talk about how to prepare (1) substance. To learn more about how to improve your interview technique for a <a href="https://medium.com/@tonygwu/mastering-the-remote-system-design-interview-de7120b9ea52">system design interview read this post</a>.</p><p>One universally recommended resource is <a href="https://www.amazon.com/Designing-Data-Intensive-Applications-Reliable-Maintainable/dp/1449373321">Designing Data-Intensive Applications</a> by Martin Kleppmann (also known as DDIA). I recommend all backend or full-stack engineers to read it at some point in their careers. In fact, I’m certain that if you: (1) read that entire book cover-to-cover, and (2) retain and conceptually understand all the material, (3) can actually apply the concepts to a real-world problem, you’ll be well-prepared for any system design interview.</p><p>However this is challenging for several reasons:</p><ol><li>The book is a fairly long, and it’s the type of non-fiction that is very technical, which makes it difficult to stay focused reading for long periods of times.</li><li>To thoroughly read it cover-to-cover and internalize all the concepts, might take you 15–60 hours (depending on your velocity and pre-existing familiarity).</li><li>Since it’s so long, as you read new material you’ll inevitably forget some of the previously read material, especially if you’re not applying it or “interacting with the material” in an interview or real-world setting.</li></ol><h3>A Targeted Approach</h3><p>Therefore, I recommend taking a targeted and pragmatic approach to preparing for the system design interview, that involves a combination of not just reading the Design Data-Intensive Applications book (aka DDIA), but also quizzing and interacting with materials/problems as you progress through it. This study technique involves the following:</p><p><strong>(A) Prioritize content </strong>(or chapters) based on likelihood it’ll be useful in a standard System Design interview.</p><p><strong>(B) Take a Diagnostic Quiz</strong> to evaluate which areas you’re weakest on, in order to dive deeper to study.</p><p><strong>(C) Read interactively</strong>. As you read through the relevant chapters of DDIA, periodically do two things:</p><ol><li><strong>Use ChatGPT</strong> to help you understand and grasp concepts in further depth.</li><li><strong>Apply learning in a practical setting </strong>by speaking through parts of a system design question (with an Excalidraw/Lucid/Google Draw) related to the area you just covered.</li></ol><p><strong>Side Note</strong>: One thing to keep in mind is that systematically preparing for a system design interview is also a great way to improve your technical skills as an engineer. So even if you put in the effort to study and improve, and it doesn’t yield an offer from a company, it will still be extremely valuable for your growth and technical development.</p><h4>Prioritize Content + Diagnostic Quiz</h4><p>I’ve identified 8 baseline topic areas to cover. Note this list might not be 100% complete, it’s a very good starting point that captures most useful areas for a system design interview:</p><ol><li>Data models, Storage, Retrieval</li><li>Interface and API Design</li><li>Replication</li><li>Partitioning/Sharding</li><li>Load Balancing</li><li>Consistency</li><li>Batch Processing Data</li><li>Streaming and Messaging</li></ol><p>To identify which of these 8 areas you need to prioritize, I’ve designed a <a href="https://docs.google.com/forms/d/e/1FAIpQLScTpcvY4En1VB0pMd9b8E5EPxC35l-XjXf1Yn0cuW5KfJlF5g/viewform?usp=sf_link">quiz to help you test yourself</a>. The quiz doesn’t require login, and will give you answers afterwards about how well you performed.</p><p>Depending on which of the 8 areas you need to devote more effor to, I recommend specific reading for you to cover. <strong>Note</strong>: DDIA refers to the book <a href="https://www.amazon.com/Designing-Data-Intensive-Applications-Reliable-Maintainable/dp/1449373321"><strong>Designing Data-Intensive Applications</strong></a> (as previously mentioned).</p><h4>If you need to focus on…</h4><h4>… then I recommend you check out….</h4><ul><li>(1) Data models, Storage, Retrieval<br>- <strong>DDIA Ch. 2–3</strong> (Data Models and Storage and Retrieval)</li><li>(2) Interface and API Design<br>- <strong>DDIA Ch. 4</strong> (Encoding and Evolution)<strong><br>- </strong><a href="https://medium.com/devglossary/what-is-apache-thrift-is-it-the-same-as-grpc-2562dba125b0">https://medium.com/devglossary/what-is-apache-thrift-is-it-the-same-as-grpc-2562dba125b0</a></li><li>(3) Replication<br>- <strong>DDIA Ch. 5</strong> (Replication)</li><li>(4) Partitioning/Sharding<br>- <strong>DDIA Ch. 6</strong> (Partitioning)</li><li>(5) Load Balancing<br>- <a href="https://medium.com/the-kickstarter/load-balancing-101-81710aa7a3d7">https://medium.com/the-kickstarter/load-balancing-101-81710aa7a3d7</a><br>- <a href="https://blog.bytebytego.com/p/ep47-common-load-balancing-algorithms">https://blog.bytebytego.com/p/ep47-common-load-balancing-algorithms</a><br>- <a href="https://www.youtube.com/watch?v=dBmxNsS3BGE">https://www.youtube.com/watch?v=dBmxNsS3BGE</a></li><li>(6) Consistency<br>- <strong>DDIA Ch. 9</strong> (Consistency and Consensus)</li><li>(7) Batch Processing Data<br>- <strong>DDIA Ch. 10</strong> (Batch Processing)</li><li>(8) Streaming and Messaging<br>- <strong>DDIA Ch. 11</strong> (Stream Processing)<br>- <a href="https://www.youtube.com/watch?v=JalUUBKdcA0">https://www.youtube.com/watch?v=JalUUBKdcA0</a></li></ul><p>By taking the quiz, it’ll short-circuit your reading time/effort and have you only focus on the materials that are most relevant to your weaknesses.</p><h4>Studying Interactively</h4><p>As you go through and review the various topics (either YouTube videos or reading DDIA or other materials), you should try to study interactively and test yourself as you’re reading/consuming the content. Don’t just read the material cover-to-cover, but also test your knowledge and engage with the content interactively, as it will both help you conceptualize and retain it for your interview. You can do so by:</p><p><strong>(1) Use ChatGPT to ask follow-up questions</strong>. Whenver you read a concept/topic that is a bit foreign, ask ChatGPT follow-up questions about it. For example, if you’re reading about quorum reads or quorum writes in DDIA and want to internalize the concept a bit more, you can ask ChatGPT:</p><ul><li>Ex: “Can you explain in more details how <strong>quorum reading</strong> works?”</li><li>Ex: “What are some examples of databases that use <strong>quorum reading</strong> to handle replication/consistency?”</li><li>Ex: “What are the disadvantages of using <strong>quorum reads</strong>?”</li><li>Ex: “Do Cassandra, HBase, RocksDB use <strong>quorum reads</strong>?”</li><li>Ex: “What are some alternative mechanisms to achieve the same outcome as <strong>quorum reads</strong>?”</li></ul><p>You can substitute “quorum reads” with any topic X that you read or learn about.</p><p><strong>(2) Practice what you’ve learned in an interview context</strong>. Take any random practice interview question (e.g. from either of Alex Xu’s books <a href="https://www.amazon.com/System-Design-Interview-insiders-Second/dp/B08CMF2CQF/ref=asc_df_B08CMF2CQF/?tag=hyprod-20&amp;linkCode=df0&amp;hvadid=459440273404&amp;hvpos=&amp;hvnetw=g&amp;hvrand=6590450694177813807&amp;hvpone=&amp;hvptwo=&amp;hvqmt=&amp;hvdev=c&amp;hvdvcmdl=&amp;hvlocint=&amp;hvlocphy=9061275&amp;hvtargid=pla-934212337151&amp;psc=1&amp;mcid=5ceb382192cb3da0872c58345ba41616&amp;gclid=CjwKCAiAqNSsBhAvEiwAn_tmxYQqFsvSEgOy7U0snEmDzsDvgP5UfoJTzF3b_oATibUZ5rwCWj87URoCtWIQAvD_BwE">volume 1</a> or <a href="https://www.amazon.com/System-Design-Interview-Insiders-Guide/dp/1736049119/ref=asc_df_1736049119/?tag=hyprod-20&amp;linkCode=df0&amp;hvadid=564675582183&amp;hvpos=&amp;hvnetw=g&amp;hvrand=6590450694177813807&amp;hvpone=&amp;hvptwo=&amp;hvqmt=&amp;hvdev=c&amp;hvdvcmdl=&amp;hvlocint=&amp;hvlocphy=9061275&amp;hvtargid=pla-1645933631661&amp;psc=1&amp;mcid=a82cf81885ab3d3583a8181e3a63a392&amp;gclid=CjwKCAiAqNSsBhAvEiwAn_tmxfT9fzad1RLkjyBlEJR7Ki5Kae4zBDq-3Nuh-7dHOl67lD_4EZ5Y9hoC5TEQAvD_BwE">volume 2</a>) and do a quick mini practice (you can video record your screen with an Excalidraw/Lucid/Google Draw/Whiteboard) talking through how the given concept you learned applies to the given interview problem.</p><p>For example, if you are reading about <strong>replication</strong>, you can practice discussing how you’ll handle that topic for any of the sample interview questions (e.g. how would you set up <strong>replication</strong> for the URL Shortener or the Notification System or the GeoSpatial Location Server or the Chat System).</p><p>In summary, you can take a more targeted and thoughtful approach in preparing for the system design interview, by (1) <a href="https://docs.google.com/forms/d/e/1FAIpQLScTpcvY4En1VB0pMd9b8E5EPxC35l-XjXf1Yn0cuW5KfJlF5g/viewform?usp=sf_link"><strong>quizzing yourself</strong></a><strong> </strong>to see where you’re weaknesses are, (2) focusing only <strong>on those relevant</strong> chapters/materials, and (3) studying <strong>“interactively”</strong> as opposed to passively.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d89b10991151" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Guide to Technical Presentations / Retrospectives]]></title>
            <link>https://tonygwu.medium.com/interview-prep-project-presentations-technical-retrospectives-4348a8467218?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/4348a8467218</guid>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[interview-questions]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Thu, 26 Oct 2023 03:51:59 GMT</pubDate>
            <atom:updated>2023-10-26T05:36:09.446Z</atom:updated>
            <content:encoded><![CDATA[<p>In this Medium Post, we’ll talk about how to prepare and do well in a “project presentation” or “technical retrospective” type interview. Typically in this scenario you’ll be asked to prepare a set of artifacts that help you present something technical that you worked on in the past. These types of interviews are called many different things (but they’re effectively very similar):</p><ol><li><strong>Technical Deep Dives or (“Reverse Pirate”)</strong>: Often you’ll be interviewed on some technical initiative of project that you led in the past. Part of this interview can be categorized as a behavioral question because they offer the interviewer the ability to probe on your prior experience and require many of the similar principles in this doc, even if they can be technical in nature.</li><li><strong>Technical/Project Presentations</strong>: Often you’ll be asked to prepare some artifact (e.g. Slides).</li></ol><h3>Selecting the technical project</h3><p>It’s important that you select a technical project that is extremely deep and can showcase your strong understanding and decision-making. The project needs to be difficult and challenging in multiple technical ways. It should be the type of project that more junior engineers can’t handle, so you can showcase your technical prowess.</p><p>Here are some technical dimensions of “difficulty”. Try to pick your technical project that is difficult along multiple of these dimensions.</p><p><strong>(1) Scale</strong>. Your technical project needed to scale (which would make it more difficult).</p><ul><li>Ex: It requires special type of sharding/partitioning, replication, etc.</li></ul><p><strong>(2) Availability/Reliability</strong>. Your project required some challenging type of availability or reliability that is not trivial.</p><p><strong>(3) Latency/Throughput</strong>. Your project required some type of special latency/throughput that made it challenging.</p><p><strong>(4) Customers/Business</strong>. Your project had some especially difficult or challenging customers/stakeholders/business requirements.</p><p><strong>(5) Security/Privacy</strong>. Your project had some security or privacy constraints that added a great deal of complexity to it.</p><p><strong>(6) Multi-modal/factorial</strong>. Your project involved many very disparate systems working together (e.g. mobile + ML + infrastructure, etc.)</p><h4>Merging/Consolidating Projects</h4><p><strong>Note</strong>: if the technical projects you’ve worked on are only complex in one of the above ways, you can combine them together to form a more complex project to present.</p><ul><li>Ex: You worked on multiple experimentation products, one that allows you to set up an experiment (aka Gatekeeper or Quick Experiment) and the other that lets you analyze an experiment (aka Deltoid).</li><li>Let’s say the first product has crazy scale (1) and availability (2) requirements, and the second had challenging stakeholder (4) and security (5), you can combine these two projects together.</li><li>You can frame the technical deep dive as building an entire experimentation product suite end-to-end that has four separate complexities (1, 2, 4 and 5).</li></ul><h4>Projects to Select vs. Avoid</h4><p>In my experience, there are a few “categories of projects” that I think generally make good projects to select, and a few that i think you should avoid.</p><p><strong>Great Projects to Select</strong>:</p><ul><li><strong>0 → 1 and then 1 → 10 Projects</strong>: These are great projects to discuss because they typically involve interesting tradeoffs.</li><li><strong>Well Known Projects</strong>: Projects whose complexity is publicly well-known are great projects to discuss. For ex: Building News Feed at Facebook, or Ethereum wallets at Coinbase, etc. These will help save you from having to explain context on the project to the interviewer, as they will likely already intuit the importance and complexity.</li><li><strong>Platform Projects</strong>: Projects on top of which other engineers need to build, provide really interesting discussion points.</li></ul><p><strong>Types of Projects to Avoid</strong>:</p><ul><li><strong>Migrations:</strong> Migrating from one system to another generally appears less impressive to an interviewer than it does to you. The reason is because the complexity of a migration usually comes from operational overhead, ensuring feature parity, and backfilling. To an interviewer without much context on the complexity of your existing system, this will not appear as impressive/complex as it really was.</li><li><strong>Simple Features</strong>: Implementing simple features within a product (e.g. add a UI component, etc.) often don’t appear as interesting for an interviewer, because for them to intuit the impact of the simple feature, they’d have to have a deeper understanding of your prior company’s product suite.</li></ul><h3>Preparing Artifacts</h3><p>You’ll likely want to prepare one or more artifacts, while also not making it appear too rehearsed. You can either create a set of slides or use a white boarding tool like Excalidraw.</p><h4>Excalidraw or Lucid</h4><p>If you’re using a product like Excalidraw or Lucid, I recommend to have a full version of your diagramming already pre-drawn out and open in a separate window during the interview. This will allow you to quickly have a “north star” that you’re diagramming towards and you can reference it during the interview.</p><h4>Slides</h4><p>If you do prepare a set of slides, keep it simple and make sure you talk about the business context first as well as the constraints/challenges of what you’re building.</p><h4>Technical Choices</h4><p>Be prepared to talk in depth about all of the key technical choices you made. If you’re not sure about a given technical decision (and aren’t able to explain it) utilize ChatGPT to find out more details as needed.</p><ul><li>Ex: If your technical project used a technology (e.g. Simple Queue Service), but you’re unsure what the alternatives (e.g. Service Bus, RabbitMQ, Kafka, etc.) were and why that was selected, you can prompt ChatGPT to learn about the details.</li><li>ChatGPT Prompt: “I’m trying to find a technology to handle queueing for my consumer web app. The existing app is on the AWS stack, what technologies can I use besides Amazon SQS?” → ChatGPT will give you some answers/thoughts.</li></ul><h3>During the Interview</h3><p>During your interview, keep a few things in mind:</p><ul><li><strong>Zoom in or out as needed. </strong>Read cues from your interviewer and be able to zoom in and out regarding the level of technical detail your interviewer wants from you. Offer to talk about things in more detail as needed.</li><li><strong>Use structure, number your lists</strong>. The similar rules apply to <a href="https://medium.com/@tonygwu/tips-for-non-functional-interviews-e2da54a03b0e">behavioral interviews</a> — try to use numbered lists and structure to navigate the conversation efficiently.</li><li><strong>Offer specific discussions as needed</strong>. Offer to go into more detail on specific areas of the project to your interviewer as needed.</li><li><strong>Terminology and Technical Decisions</strong>. Be prepared to explain any terminology that arises as well as technical decisions. If you say your project used X, be prepared to explain X and why did you utilize X.</li><li><strong>Time Management</strong>. Keep an eye on the clock to make sure you have enough remaining time to cover more interesting aspects of the interview.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4348a8467218" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Behavioral Interviews: Stories (TMW)]]></title>
            <link>https://tonygwu.medium.com/behavioral-interviews-stories-tmw-6a61c41292c2?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/6a61c41292c2</guid>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[interview-questions]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Thu, 26 Oct 2023 03:41:19 GMT</pubDate>
            <atom:updated>2023-10-26T04:45:12.188Z</atom:updated>
            <content:encoded><![CDATA[<p>As part of <a href="https://medium.com/@tonygwu/tips-for-non-functional-interviews-e2da54a03b0e">behavioral interviews</a>, you’ll often be faced with a “Tell Me When” or “Give an example” interview question. I call these questions TMW or GAE questions. The purpose of this post is to give you a definitive guide for how to prepare and successfully answer these types of questions.</p><h3>Recognizing it’s a TMW/GAE question:</h3><p>I generally define a TMW/GAE question as one where your interviewer expects you to tell them a story. The most obviously way to recognize it as a TMW or GAE question based on the phrasing of the question:</p><ul><li>Ex: “<strong>Tell me when</strong> you had to manage tough expectations with a peer.”</li><li>Ex: “<strong>Give an example</strong> where you changed your manager’s mind about something significant.”</li></ul><p>However, there might be questions for which it’s subtle or ambiguous whether your interviewer expects a story:</p><ul><li>Ex: “When did you last have a disagreement with your PM partner?”</li><li>Ex: “What do you do to manage low performers on your team?”</li></ul><p>If it’s unclear if the interviewer is asking you a TMW/GAE question (and thus expecting a story), you should immediately align with the interviewer really quickly. You can do this by asking a clarifying question, as an example:</p><ul><li>Interviewer asks: “What do you do to manage low performers on your team?”</li><li>Answer: “I’ve done performance management for a lot of low performers, are you looking for me to give some general principles/framework I use, or <strong>do you want me to give a specific example of when I handled low performers on my team?</strong>”</li></ul><p>If you have any ambiguity at all, asking a clarifying question will quickly help you align to ensure you are answering a GAE/TMW question, and thus the interviewer expects a story from you.</p><h3>Preparing Stories</h3><p>You’ll need <strong>prepare 8–15 stories ahead</strong> of time to cover these TMW/GAE questions.</p><h4>Stories are really compelling</h4><p>If the question is a “Tell Me When” question, you should expect to be able to tell a story. However, even in some cases when the interviewer asks you a “How do you” question, it can often be compelling to tell a story in order to supplement your answer.</p><p>Commonly during the interview you’ll be asked situational questions (e.g. “have you ever had to manage someone out” or “have you ever had to resolve a conflict”). The interviewer is looking for to see the breadth and depth of your experience within the specific domain.</p><p>Generally speaking, after you’ve ensured the interview is a TMW/GAE question, you should now pivot to telling a story rather than merely answering a question.</p><ul><li><em>For Example:</em> If the interviewer asks you a question like “how do you generally handle stress of project deliverables”, instead of answering the question at face value, probe the interviewer to see if they’d also like you to answer the question with a story.</li><li>Ex: “That’s how I generally handle stress when I have a hard deadline…, let me give you an example of my doing this in practice?”</li></ul><p>Practice both telling your stories fluidly as well as transitioning from the questions into the story — you can do this by video recording yourself and playing it back.</p><h4>Picking Stories</h4><p>In order to prepare for behavioral interviews, you should prepare around 8–15 stories that you can draw upon at will. Generally each story you prepare, likely can be used for a wide variety of potential interview questions.</p><p>Try to pick stories that demonstrate multiple values or skills that you have. The best stories can be used to showcase 5–6 different positive attributes about yourself.</p><p>For example, let’s say you have a story about how your manager and you disagreed on how to help a colleague who had communication challenges — you worked a few nights to improve her communication and presentation skills so she could demo her project. This story is useful for a variety of different interview questions:</p><ul><li>ex: What was a time in which you helped a colleagues?</li><li>ex: How do you handle others with weaknesses?</li><li>ex: How do you handle a problem that is ambiguous or ill-defined?</li><li>ex: Describe how you disagreed with a team or manager decision?</li></ul><p>Think through the entire space of questions that could be asked of you (questions about your mindset, questions about your experience/behavior, questions about your motivation).</p><p>Even answers to questions that seem straightforward (e.g. what are you looking for in your next role?) can be made to seem more authentic if you connect with a story (e.g. tell about how you decided on your current role and what did you learn).</p><h4>Difficulty + Leveling</h4><p>As you pick the story, ensure that the story is as difficult/challenging as possible (at least commensurate with the level you’re interviewing for). If you’re looking for a very senior leadership position, your story can’t just be about fixing some code reviews. Remember, the more difficult the story the better.</p><p>Additional resource: <a href="https://www.youtube.com/watch?v=hU6BVxtGd5g">Video about U-shaped story-telling</a>.</p><h4>Story Format: STAR</h4><p>Must has already been written about the <a href="https://www.vawizard.org/wiz-pdf/STAR_Method_Interviews.pdf">STAR method</a> for coming up with a story, so I won’t go into much further detail, other than to highlight a few additional points.</p><p><strong>(1) Don’t be afraid to be explicit</strong>. This might sound awkward, but it is actually extremely helpful to be explicit with your interviewer when moving from each stage to each.</p><p>Ex: “Let me start off with the <strong>situation</strong> I faced...”</p><p>Ex: “Now I’ll tell you about the <strong>problem/task</strong> I was responsible for…”</p><p>Doing so allows your interviewer to follow along and makes it much easier for them to take notes.</p><p><strong>(2) Be prepared to still be flexible</strong>. Despite using the STAR method, still be prepared to be flexible with your interviewer. Your interviewer might want you to skip one or more of the sections and jump straight to the result, or they might just want to skip all of the S, T, A, R sections altogether. Don’t just blindly tell the STAR story without regards to what your interviewer is asking.</p><p><strong>(3) Use the word “I” instead of “we” when describing T and A</strong>. The interviewer wants to know what you did and what you were responsible for, not the team.</p><p><strong>(4) Flexibly offer additional letters/sections</strong>. Where appropriate, flexibly offer to the interviewer additional components to your story besides (S), (T), (A), (R). For example, you can also offer to tell the interviewer about your <strong>(L) learnings</strong> from the story: “I can also tell you a little bit more about my learnings and how I’d handle the situation differently in the future.”.</p><ul><li><strong>(M) Mistakes</strong>: You can offer to include what you think are mistakes (e.g. “happy to also tell you about the mistakes I made in this process”).</li><li><strong>(L) Learnings</strong>: You can also offer to share what your learnings are from this (e.g. “I learned a lot from this process, happy to share what those learnings are).</li><li><strong>(C) Additional context</strong>: You can also offer to describe additional context besides (S) in more depth (e.g. “Given this situation, I can also provide a bit more context about XYZ).</li><li><strong>(O) Obstacles/Challenges</strong>: It might make sense to offer to the interviewer additional obstacles (e.g. “To complete this project/task, I also faced 3 additional challenges: X, Y, Z. Happy to go more into depth about these obstacles”).</li></ul><h4>Merging / Consolidating Stories</h4><p>You likely have more than 15+ total stories you can draw from, but it’s helpful for your own sanity to merge/consolidate them together. There’s many different types of merging that you can do:</p><ol><li><strong>Common characters</strong>. If you have multiple stories centering around the same person (for ex: the product manager on your team or your manager), you can probably consolidate them into a singular story.</li><li><strong>Similar actions</strong>. If you took similar actions (for ex: you refactored your team’s codebase in multiple instances), you can likely consolidate them into a singular story to tell.</li><li><strong>Similar timeline/contexts</strong>. For example, let’s say during the US 2020 election at Facebook to combat misinformation, you had to take on multiple heroic efforts separately that spanned 1–2 months. You can probably consolidate this into a singular story (with shared context) that you can tell.</li></ol><h4>Different Length Stories</h4><p>Now that you’ve prepared the stories, you should practice telling them effectively. I recommend you prepare different length stories. It might be good to have a <strong>short</strong> (30 seconds), <strong>medium</strong> (2 minutes) and <strong>long</strong> (6–7 minutes) version of the stories you’d like to tell.</p><p><strong>Example short answer/story</strong>: “I hired an engineer on our team, and I knew he wanted to do a startup in the long-run. He stayed on the team and we helped him grow from an IC3 to IC5. His interests aligned with the team when he was junior, but once he became a senior engineer, the skills he would learn/develop no longer aligned with his long-term startup desires, so I helped him find a new team in the Growth org. We used the headcount to hire scalability and infra experts to join our team, and it worked out best for everyone. I learned the importance of keeping in mind engineers long-term career desires when planning their projects/goals.”</p><p>When you answer a question you can start off with the short-version, but be prepared to go down deeper and give the medium or long-versions if your interviewer is interested. You can proactively offer this up during or after you give your short answer with something like “<strong>Throughout this process, I learned XXX. I’m also happy to talk a bit more about what techniques I used to uncover his long-term startup affinity or how I went about finding his new team for him.</strong>” Your interviewer might decide to move onto the next question with “no that’s okay, let’s move on”, or he/she night decide to oblige and ask further questions. Even if your interviewer (for interests of time) doesn’t decide to go deep, it will still come across that your answers have a significant amount of depth, and you’ve done a good job reflecting on your experiences.</p><h3>Practice Telling the Stories</h3><p>Once you’ve written out the stories, here are afew tips for practicing them.</p><ol><li><strong>Record yourself telling the stories</strong>. It’s helpful to practice telling the stories in front of a webcam and record yourself. After watching it back, you’ll understand and capture a lot of learnings and improvements (e.g. using filler words, or being too fast/slow/rambly).</li><li><strong>Have a practice partner probe</strong>. Have a practice partner probe your story with follow-up questions to help test whether you’ve thoroughly prepared your story.</li><li><strong>Track your time-management</strong>. Track how much time you’re spending on the story.</li><li><strong>Practice offering follow-ups/clarifications</strong>. Practice pausing at various points during your story-telling and offering additional follow-ups or clarifications for your interviewer (e.g. “Let me pause, do you want me to go into more detail and talk about why these problems were especially challenging?”).</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6a61c41292c2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[General Interview Preparation]]></title>
            <link>https://tonygwu.medium.com/general-interview-preparation-494788cba1af?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/494788cba1af</guid>
            <category><![CDATA[interview-tips]]></category>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[interview-questions]]></category>
            <category><![CDATA[interview-prep]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Tue, 24 Oct 2023 04:03:07 GMT</pubDate>
            <atom:updated>2023-10-26T03:47:30.739Z</atom:updated>
            <content:encoded><![CDATA[<p>In this Medium Post, we’ll describe general interview preparation techniques for the tech industry, that apply to all interviews (coding, system design, behavioral, etc.). We’ll cover 4 specific areas: (A) planning &amp; scheduling logistics, (B) self-studying, (C) practice interviews (D) day before and day of. Finally I’ll include additional medium posts I’ve written regarding preparation for specific interview types (e.g. frontend coding, backend coding, etc.).</p><p><strong>Note</strong>: this guide is intended for interviews that are scheduled over VideoChat, thought some of it may apply to in-person over-the-table interviews as well.</p><h3>Planning &amp; Scheduling Logistics</h3><p>Depending on how well you’re prepared, you’ll want to optimize how you plan for and schedule your interviews.</p><h4><strong>Day of the Week</strong></h4><p>Pick a day of the week that you tend to do better on interviews.</p><h4><strong>Time of the Day</strong></h4><p>Some people are more alert in the morning vs. the afternoon. Select a time of the day that optimizes for your own performance.</p><h4><strong>Split across multiple days</strong></h4><p>If you have a full 5+ hours of interviews, I highly recommend that you split them across multiple days, for two reasons:</p><ol><li><strong>Optimized for your preferred time-of-day</strong>. If you’re a morning or afternoon person, you’ll be able to have all your interviews during that time of day.</li><li><strong>Combat interview fatigue</strong>. Interview fatigue is a real problem, especially if you stack 5–6 hours of interviews back-to-back.</li></ol><h4>Block of time around interview</h4><p>I recommend that you block off the the 45 minutes right before (for warmup) and 30 minutes after (reflect + over) each of your interviews. This means that you shouldn’t have interviews that go back-to-back.</p><p><strong>(A) Block off 45 min before each interview</strong>. Having 45 min. blocked off on your calendar before the interview starts is extremely helpful. For example, when listing your availability, if you’re actually free at 10:00a, tell your coordinator you’re free starting at 10:45a (so you have 45 min.).</p><ol><li>If you have an existing day job with meetings — those might run late or be particularly intense. Having a solid buffer in between a meeting from your day job and your first interview is helpful to unwind, decompress and change your frame of mind.</li><li>You can use the bathroom, double-check your room’s lighting and figure things out.</li><li>This will give you a chance to do a few interview warmups (see below on warmup section).</li></ol><p><strong>(B) Interview Warmups</strong>. It’s helpful to warmup for the first 20–30 min. right before your interview. This will help you remove pre-interview jitters and get your in the zone. To do a warmup, do a 1–2 practice questions or “question parts” for the interview by yourself in the same videoconferencing software that you’ll be using for the interview (Zoom, Google Meet, or Cisco Webex) so you can simulate the actual interview.</p><ol><li>If it’s a coding interview, do a quick practice coding problem in the same online IDE (e.g. Coderpad or HackerRank) that your interview will be in.</li><li>If it’s a system design interview, practice talking through the first several minutes of a sample system design problem.</li><li>If it’s a behavioral interview, find 1–2 random interview question and practice physically talking through your answer to those questions.</li></ol><p>Just as Steph Curry takes practice shots and Patrick Mahomes makes practice throws before each game, it’s likewise important for you to get your practice reps in before the interview.</p><p><strong>(C) Write out custom questions for interviewer</strong>. Take the time to write out 3–4 custom questions for the interviewer. The questions should be custom to <strong>(1) the company, (2) the role</strong> <strong>and (3) the interviewer</strong>. You have no excuse for not asking an extremely intelligent question of the interview, since this is something you can think about a long time before the interview.</p><ol><li>If it’s a question you can find out by searching online (ex: “when was the company’s last round of funding”) <strong>it’s a bad question</strong>.</li><li>If it’s a question that could apply to any generic company (e.g. “how does the performance process work?”) <strong>it’s a bad question</strong>.</li></ol><p>For more tips on preparing the right set of questions to ask your interviewer, <a href="https://tonygwu.medium.com/questions-for-your-interviewer-e64228dbe225">read this longer post</a>.</p><p><strong>(D) Block off 30 min after</strong>. Ask your recruiting coordinator or scheduler if you can have 30 minutes break in between interviews if possible. This allows you decompress after the prior interview, use the bathroom and take a break, as well as reflect. Some candidates use this time to meditate, especially if the previous interview was not as successful as you would have liked.</p><h3>How to self-studying</h3><p>A good chunk of your interview preparation will involve self-studying and preparing for the interview on your own. In this section, we’ll include a few general tips for doing the self-studying.</p><p><strong><em>Note</em></strong><em>: at the bottom of this post, I’ll include links for how to prepare for specific interview types (coding, system design, behavioral, etc.).</em></p><p><strong>(1) Understanding Each Interview</strong>. Ask the recruiter (or any friends you know at the company) what each of the interviews will be like. Often a recruiter will chat with you on the phone about each of the types of interviews and answer any questions you have. <strong>Remember</strong>: All things being equal, your recruiter is in your corner and trying to help you get an offer.</p><ol><li>See if your recruiter can verbally discuss what the rubric for the interview will look like (ie. what does a “Strong Hire” vs. a “Weak Hire” look like?</li><li>Ask the recruiter what the common ways candidates have failed the given interview. Your recruiter might tell you something like: “on the coding, candidates often run out of time”, or “for system design, candidates often make their solutions too complicated”. Having this insight will be helpful in your preparation and you can prepare accordingly.</li></ol><p><strong>(2) Prioritize what to study</strong>. You likely have a really limited amount of time to prepare and study, so don’t try to practice everything. Based on your conversations with the recruiter, focus your preparation on a few key areas. I’ve seen ambitious candidates try to do too much (for ex: they did 600+ Leetcode problems and <a href="https://www.amazon.com/Designing-Data-Intensive-Applications-Reliable-Maintainable/dp/1449373321">Designing Data Intensive Applications</a> cover-to-cover), and then fail to get to all of it, and as a result failed to prep the highest probability areas.</p><p><strong>(3) Don’t forget to practice soft interview skills</strong>. Even for technical or functional interviews, don’t forget to continue to practice your soft interview skills<strong>.</strong></p><ol><li><strong>Identify clarifying pausing</strong>. For each interview question you’re practicing (whether it’s coding, design or behavioral) practice thinking about and verbalizing which clarifying questions you’d ask for the given question.</li><li><strong>Develop habit of pausing and aligning</strong>. As you do an interview, practice pausing part-way through an interview and checking with the interviewer. “Let me pause here and see if you’d like me to go deeper into this area, or move on?”</li></ol><h3>Doing Practice Interviews</h3><p>When doing practice interviews (with a practice partner), here are some general tips:</p><ol><li><strong>Simulate the real interview setting as much as possible</strong>. This means having the practice interviewer time the interview and use the same time controls as you would in the real interview. This also means using the same tooling (HackerRank, Excalidraw, CoderPad, etc.) that you would in the real interview setting.</li><li><strong>Select a strong practice partner</strong>. You’ll want to select a practice partner that is stronger than you are, someone who can give you critical feedback to help you improve.</li><li><strong>Record the sessions and watch them back</strong>. Obviously only do this with permission from your practice partner. Recording the session and watching it back is one of the best ways to learn. There are tools you can use to record your computer screen as well as your audio at the same time.</li><li><strong>Generalize learnings, and develop habits</strong>. Every time you practice and learn things from the practice session, see if you can generalize those learnings to not just apply to the given interview question or problem. One way to codify the learnings is to try to develop general habits to follow. For example, say you run out of time during an interview session. Instead of simply concluding that you need to get through the problem faster, you could generalize that you need to get into the habit of asking the interviewer how many parts the interview will be, this will help you keep pace with the interview.</li></ol><h3>More Reading</h3><ol><li><a href="https://tonygwu.medium.com/frontend-coding-interview-prep-42243842bafd">Frontend Coding Interviews</a>. How to prepare for a frontend coding interview.</li><li><a href="https://medium.com/@tonygwu/coding-interview-preparation-07ee75fd3753">Backend Coding Interviews</a>. How to prepare for a backend coding interview.</li><li><a href="https://medium.com/@tonygwu/mastering-the-remote-system-design-interview-de7120b9ea52">System Design Interviews</a>. How to prepare for a system design interview.</li><li><a href="https://medium.com/@tonygwu/tips-for-non-functional-interviews-e2da54a03b0e">Behavioral Interviews</a>. How to genenearlly prepare for a behavioral (or non-functional) interviews.</li><li><a href="https://tonygwu.medium.com/behavioral-interviews-stories-tmw-6a61c41292c2">Behavioral Interviews: Stories</a>. How to prepare, practice, and tell stories to tell during a behavioral interview.</li><li><a href="https://medium.com/@tonygwu/questions-for-your-interviewer-e64228dbe225">Asking Questions of your Interviewer</a>. How to select questions to ask your interviewer.</li></ol><h3>More Resources:</h3><p><a href="https://www.youtube.com/watch?v=Sb0p1-TGjmc">Learnings from Tech Interviews at Amazon</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=494788cba1af" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Questions for your Interviewer]]></title>
            <link>https://tonygwu.medium.com/questions-for-your-interviewer-e64228dbe225?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/e64228dbe225</guid>
            <category><![CDATA[interview-questions]]></category>
            <category><![CDATA[interview-tips]]></category>
            <category><![CDATA[interview-prep]]></category>
            <category><![CDATA[interview]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Tue, 24 Oct 2023 03:22:13 GMT</pubDate>
            <atom:updated>2023-10-24T03:22:13.304Z</atom:updated>
            <content:encoded><![CDATA[<p>During any interview, your interviewer will almost certainly leave 5–15 minutes at the end of the interview for you to ask them questions (ie. “do you have any question for me?”) about the company or role.</p><p>On one hand, this won’t be the most important signal in determining if you get an offer, but on the other hand, it’s extremely easy to prep for, so there’s no excuse for you not to have a great set of questions for your interviewer. I’ve seen so many candidates have a set of bad or mediocre questions prepared, that I think it’s worth discussing here how to come up with a really amazing set of questions.</p><h4>Custom to the Company</h4><p>The questions should be<strong> custom to the company</strong>. If you ask a question that could easily apply to any other company, then that’s generally a bad question</p><p><strong>Example</strong>: “Can you tell me what the culture is like at Facebook?”.</p><ul><li>This is a bad question because it can be applied to any company. You’re not signaling anything meaningful about your interest or knowledge of the company because you could turn around an ask the same question in your next interview.</li></ul><p><strong>Example</strong>: “Given Facebook has grown to be a large scale thus far with 50k+ employees, how do you still maintain the pace of velocity?”</p><ul><li>This is a slightly better question, in that you could only ask this question for a handful of companies (e.g. FAANG companies).</li></ul><p><strong>Example</strong>: “What has been the strategy for incorporating the Metaverse into Facebook’s existing Family of Apps like Instagram, News Feed, WhatsApp and Messenger given those apps follows a feed experience, while the Metaverse follows a 3-D immersive experience?”</p><ul><li>This is a strong question, as there’s literally only one company in the world for which you could ask this question.</li></ul><h4>Custom to the role</h4><p>The questions should be<strong> custom to the role</strong>. Even if your question is company-specific, it should ideally be targeted to the role you’re interviewing for.</p><p><strong>Example</strong>: “How much latitude do engineers have when it comes to defining what is on their roadmap at Facebook?”</p><ul><li>This is not a great question because, almost any candidate could ask this question.</li></ul><p><strong>Example</strong>: “Since I’m interviewing for an ML Platform role, I’d love to know how much I can use my personal experience using PyTorch to help prioritize what we work on?”</p><ul><li>This is a better question because it narrows the custom number of teams/roles for that this question could apply to. However the best question for an interviewer might be:</li></ul><p><strong>Example</strong>: “I follow the PyTorch subReddit and they’ve mentioned a lot of open source community requests aren’t prioritized as much as Facebook internal feature needs. If i were to join the PyTorch Infra team, can I put together user feedback panels from some of the community’s power users for the team?</p><ul><li>This question is obviously better than the previous two, as it’s extremely custom to the company, team and role.</li></ul><h4>Custom to the interviewer</h4><p>The recruiters should have give you a heads up around who will be on your interview panel, and you should think about how you can make your question custom for that specific interviewer. Imagine you’re interviewing for Instagram Ads Platform team, and one of the interviewers on your panel is a manager for the overall Facebook Ads Platform team of which your team is a customer.</p><p><strong>Example</strong>: “How does collaboration generally work between a customer and platform team at Facebook? Do I inform your roadmap?”</p><ul><li>This question is not great because it doesn’t leverage anything specific about the interviewer’s role.</li></ul><p><strong>Example</strong>: “What’s are your team’s top priorities when it comes building a robust ads platform?”</p><ul><li>This question is somewhat better, but it still can apply to other ads-related roles that you’re interview for, it’s not the most ideal.</li></ul><p><strong>Example</strong>: “I read that Facebook has an auction system where you predict the clickthrough probability of an ad and combine that with a bid value to rank ads. Does Instagram’s ad system extend that auction system, or did Instagram build its own? Since Instagram has a larger proportion of brand advertising, and less action/click-based did they build their own infra?”</p><ul><li>This is a much better question, as it makes it clear you know what you’re talking about, and it’s a question that is custom to the given interviewer. It’s a question likely that interviewer would be able to answer (and nobody else).</li></ul><h4>Using chatGPT to generate questions</h4><p>If you have difficulty coming up with custom interview questions for the given role and interviewer. For example, you can ask the following prompt of ChatGPT:</p><pre>I&#39;m interviewing to lead the Machine Learning Platform team at Waymo,<br>a self-driving car company. My interviewer is the Director of Engineering<br>who leads the Sensing and Perception team, and his team will be a customer<br>of my team. I want to ask some questions at the end of the interview that<br>are very custom to the interviewer and role. What are some good questions<br>to ask that really demonstrate I have self-driving car domain knowledge and<br>have done my research on the company?</pre><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e64228dbe225" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Frontend Coding Interview Prep]]></title>
            <link>https://tonygwu.medium.com/frontend-coding-interview-prep-42243842bafd?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/42243842bafd</guid>
            <category><![CDATA[interview-tips]]></category>
            <category><![CDATA[coding-interviews]]></category>
            <category><![CDATA[interview-questions]]></category>
            <category><![CDATA[interview-preparation]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Sat, 21 Oct 2023 16:55:04 GMT</pubDate>
            <atom:updated>2023-10-21T16:55:04.397Z</atom:updated>
            <content:encoded><![CDATA[<p>In this Medium post we’ll talk about how to prepare for a coding interview if you’re a frontend engineer at any large FAANG company. Note, this post assumes you’ll be using React and Javascript/Typescript as the core technologies you’ll use, but the principles will apply if you’re using a different set of frontend technologies.</p><p><em>Here’s the </em><a href="https://tonygwu.medium.com/coding-interview-preparation-07ee75fd3753"><em>post if you’re preparing for a coding interview as a backend engineer</em></a><em>.</em></p><h3>Overview of different classes of skills</h3><p>Learning to be great at frontend interviews is a combination of three different classes of skills, and you should aim to practice each one individually, as they come together to form a strong interview performance:</p><p><strong>(1) Problem Understanding &amp; Expectations Alignment</strong>: This comes down to how well you do understand the problem and have aligned expectations with the interviewer.</p><ul><li>One of the biggest mistakes you can make during the interview is to building something that interviewer is not expecting.</li><li>Once you’re presented with the problem, there is an <strong>art/skill</strong> to asking follow-up questions that give you valuable information about what to build, while also not appearing as “<em>asking for too many hints</em>”.</li></ul><p><strong>(2) Problem Solving &amp; Communicating</strong>: It’s important as you make progress on the problem to communicate what is happening.</p><ul><li>Interviewers will give you credit and maybe hints if you are talking about your progress as you work on the problem.</li><li>It’s important to have an intuition for what parts of the problem to tackle first (ie. functionality as opposed to the UI and styling).</li><li>A good approach to sequencing might be something like: (1) Writing out basic structure of components, (2) Implementing core functionality, (3) Styling, (4) Refinement, (5) Testing. After each stage check-in with the interviewer and ensure you’re on the right track.</li><li>Alternatively, if there are multiple milestones or functionalities that you can build incrementally and test them as you’re working on the functionality.</li></ul><p><strong>(3) Debugging &amp; Testing</strong>: your code will not work immediately when you write it. How quickly you can debug and fix issues will be important.</p><h4>Skill (1): Problem Understanding/Expectations Alignment</h4><p>You should practice asking the right questions to make sure you’re building the right things in an interview setting. Find a partner and practice asking each other questions depending on the practice interview question.</p><p><strong>(1) Ask interview meta-questions</strong>. It’s always good to ask the interviewer a set of “meta-questions” about the interview itself. Some examples:</p><ul><li>How many interview questions or parts will there be?</li><li>For each part to this question, should I look to spend equal time on each or will part 1 take more time than part 2, etc.?</li></ul><p>Understanding this can help give you understand the interviewer’s expectations (which includes how advanced/complex to make each part of the application you’re building).</p><p><strong>(2) Clarify the specific functionality expectations</strong>. Here are some example questions of this:</p><ul><li>Should I make my application mobile-friendly? Should it be responsive?</li><li>If X drop-down gets selected, should I automatically re-render Y component? Or only when the user presses Z button?</li><li>Do I need to worry about various browsers? Or I can assume it’s on Chrome?</li></ul><p>Usually “functionality” questions are great to ask since they give you signal about what the interviewer would like to see, but also don’t count as “hints” from the interviewer.</p><p>You can also demonstrate your own product intuition by asking “leading” questions, in that you have your own product opinion for how the application should work but are confirming it with th einterviewer.</p><ul><li>For example: “<em>I feel like if the user selects X dropdown, the UI should automatically re-render, since that would be the best experience, but I wanted to double-check with you.</em>”</li></ul><p><strong>(3) Clarify the tooling and technological choices you should use.</strong></p><ul><li>Will this be a large scale application or relatively limited? Should I use Flexbox or Float Layouts or Grid Layouts?</li><li>What type of performance characteristics are needed? Will I be loading a lot of data and be pre-fetching?</li><li>Will this application be write-heavy or read-heavy? What will the usage pattern look like?</li><li>Are certain workflows in this application more frequent than others?</li></ul><p>Asking these types of questions will both showcase your experience as an frontend applications developer, as well as ensure you build something that matches the interviewer’s expectations.</p><h4>Skill (2) Problem Solving &amp; Communication</h4><p>It’s important to practice problem-solving as well as communicating as you solve the problem.</p><ol><li><strong>Explain what decisions you’re making</strong>. It’s good to tie them back to the original functional constraints of the problem. “You mentioned the application will be very large scale, so I’ve opted for CSS Grid layout”.</li><li><strong>Focus on getting a basic app working first</strong>. Try to get a basic version of the application working with basic core functionality before tackling advanced capabilities.</li><li><strong>Break down functionality</strong>. Try to break functionality down into smaller bits. If there’s a piece of functionality or workflow that needs to be written, abstract it as a separate functionality so you can iterate/test it quickly.</li></ol><h4>Skill (3) Debugging and Testing</h4><ol><li><strong>Shorten the Implement → Feedback Loop</strong>. Temporarily alter your code to shorten the implementation → feedback loop. For example, if you test logic that occurs when the user clicks through a few buttons, you can hack the code to trigger that logic / call the function as soon as the page renders. This allows you to test/debug much faster, as opposed to always manually simulating the user clicks to trigger to the logical flow you’re testing.</li><li><strong>Utilize Console Logging</strong>. Be familiar with where console.log outputs in the interview IDE you’re using (e.g. CoderPad or HackerRank)</li><li><strong>Error Messages</strong>. Be familiar with all common React / JS error messages and what they potentially mean, so that you can quickly diagnose the problem. This can save you from subtle bugs that might otherwise be hard to diagnose.</li><li><strong>Live Reloading</strong>. Be familiar with the interview IDE and whether it supports live reloading so you can see changes immediately as soon as you save.</li><li><strong>Test Incrementally</strong>. Test each piece of functionality incrementally so you can catch issues earlier.</li><li><strong>Simplicity</strong>. Keep things simple and avoid premature optimization or complexity. This allows you to debug and reason about the challenges quickly. You can always mention to the interviewer, “I’ll refactor this later” or “I can probably optimize this some more after I get this working.”</li></ol><h3>How to prepare and practice</h3><p>Now that we’ve laid out the different types of skills: (1) Problem Understanding / Expectations Aligning, (2) Problem Solving &amp; Communication, and (3) Debugging &amp; Testing, let’s talk about how you should prepare and practice for your interviews.</p><h4>Get familiar with the interview IDE</h4><p>Ask your recruiter which interview tool you’ll be using (for ex: HackerRank or CoderPad), and practice your interviews with that IDE tool. Each of these tools have many different aspects, so get comfortable with using various aspects of them.</p><ol><li><strong>Console and Error Output</strong>. Where do console.log or error messages show up? Are they visible in a place that allows you to easily see and debug?</li><li><strong>Linting and Typing</strong>. When you’re using PropTypes or TypeScript how does the IDE underscore or highlight the types and check type compatibility?</li><li><strong>Running the Code</strong>. Be familiar with how to run your code. Do you have to physically press a “run” button or does it automatically reload when files get saved? How are the files (css, jsx, html) linked together in the app?</li></ol><h4>Study Core Technologies</h4><p><strong>(1) HTML and DOM/Browser Environment</strong>. Have familiarity with how the DOM works.</p><ul><li><strong>HTML5 Tags</strong>. Explore new HTML5 tags, multimedia features, accessbility attributes.</li><li><strong>DOM Manipulating</strong>. Practice manipulating the DOM using plain Javascript, understanding the rendering process and how the browser updates the DOM.</li></ul><p><strong>(2) CSS</strong>. You’re probably familiar with the basics, but spend some time understanding the more advanced parts of CSS.</p><ul><li><strong>Layouts</strong>. Practice using the different layouts (Float, Positioning, Flexbox, Grid Layout) and when to use them in different scenarios. Practice talking aloud about why you’d use one layout vs. another given a problem.</li><li><strong>Multi-Column Layouts</strong>. Also understand different multi-column layouts (column-count and column-width properties.</li><li><strong>Responsive Design</strong>. Be familiar with different mechanisms for making your web page adapt to different screen sizes (e.g. Media Queries, Fluid Layouts, or Flexbox and Grid layouts).</li><li><strong>Transitions and Animations</strong>. Have some familiarity with transitions and animations to enhance the user experience if needed.</li><li><strong>Sizing and Absolute vs. Relative Units</strong>. Understand using absolute (px, pt) vs. relative units (em, %, rem, vw, ch, etc.)</li></ul><p><strong>(3) Javascript/Typescript</strong> Get familiar with some basic javascript concepts.</p><ul><li><strong>Variables, scope, “hoisting”, closures, data types, type coercion</strong>. Be familiar with how variables and constants are defined and how scoping of these variables work.</li><li><strong>Functions, Callbacks, promises, event loop, async programming</strong>. Be familiar with functions including the “Arrow Function” syntax so that you can easily/quickly write functions and pass them around. Also get familiar with async functions and non-blocking behavior. Be familar with things like async/await.</li><li><strong>DOM Manipulation</strong>. Be comfortable with DOM manipulation.</li><li><strong>Fetching data</strong>. Get familiar with the different ways of fetching data from the server or a local file. How to use GraphQL and using RESTful APIs.</li></ul><p>(4) <strong>React. </strong>Be familiar with using React and practice building out various applications.</p><ul><li><strong>Components, States &amp; Props</strong>. Understand how to use class and functional components and manage state and pass variables via props.</li><li><strong>Lifecycle Methods</strong>: Understand the component lifecycle methods like componentDidMount, componentDidUpdate, componentWillUnmount.</li><li><strong>State Management</strong>. Be familiar with additional libraries like Redux, Zustand, Flux, ContextAPI and how to use them to manage state.</li><li><strong>Event Handling</strong>. Handle user events like clicks, form submissions and keyboard events.</li><li><strong>React Router</strong>. Practice using React Router and understand how Route, Link, and Switch are used and practice creating new routes in your application.</li><li><strong>JSX</strong>. Get familiar with JSX syntax and practice making apps with it.</li><li><strong>Data Fetching</strong>. Practice handling HTTP/API requests within React components, or loading data from the local filesystem to get your application to work.</li><li><strong>Deployment and Updates</strong>. Know ahead of time which IDE (e.g. HackerRank, CoderPad) and how it updates when you update your React app.</li></ul><h4>Practice your communication and soft skills</h4><p>When you do your mock interviews (with a practice partner), don’t just practice writing code and solving a given problem, but also practice communicating:</p><ul><li>Practice restating the problem and making sure you align with your practice partner on interview expectations.</li><li>Remember/get in a habit of asking the interviewer the interview time layout. For ex: “how many parts to the problem are there” and “will we spend equal time on each part or will we spend more for part 1 than part 2?”.</li><li>Get comfortable talking through your decision-making and thought process when it comes to making technical choices (e.g. “I’m going to make this variable a const, because xyz”).</li><li>Practice asking clarifying questions to your interview partner and have have them evaluate and recommend additional questions you could have asked during the practice session (e.g. “you could have also clarified the functionality at this point in the interview”).</li><li>When you make a mistake during the mock interview, practice verbally recovering quickly and ensuring your follow-up is correct. For ex: “I made a mistake here, I must have misunderstood the purpose of this boilerplate function, does this look better?”</li></ul><h3>Tips to Keep in Mind During Interview</h3><p>Here are a few additional things you should keep in mind as you work through your frontend coding challenge during the interview. You should pay attention to these</p><ol><li><strong>Get in the habit of asking clarifying questions.</strong> Continue to ask clarifying questions both at the beginning and throughout the interview to fully understand what’s expected of you.</li><li><strong>Mentioning improvements will give you credit even if you don’t implement them</strong>. Beyond building a base solution/application, even mentioning things will yield you credit. For example, saying something like “In the long-run I’d totally replace this css with a framework like less or sass” or “The way we fetch data here is really suboptimal, I’d replace it with X or Y”, will often give you credit with the interviewer, even if you don’t implement your own suggestions.</li><li><strong>Practice and exercise Time Management</strong>. You should always have a sense of how many “parts” exist for the interview and how much to allocate to each “part”. This will give you intuition for whether you should move faster/onto the next “part” or take your time in improving the current “part”.</li><li><strong>Pay attention to the interviewer cues</strong>. If an interviewer tells you to focus more on the current part of to move to the next part, you should follow the interviewer instructions. Remember, the goal is not to get the ideal working solution you want, but to provide the interviewer the signal needed. If an interviewer wants you to move on to the next “part”, it’s because there’s some specific signal they’re trying to gather.</li><li><strong>Don’t forget product aesthetics and usability</strong>: It’s likely the interviewer may also want to test your product intuition and usability, so it’s good to keep that in mind during the interview. Even if you end up building a solution that lacks aesthetic or usability appeal, even mentioning it (ex: “We could probably replace this massive dropdown with a search typeahead”) will be very helpful during the interview.</li></ol><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=42243842bafd" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Backend Coding Interview Prep]]></title>
            <link>https://tonygwu.medium.com/coding-interview-preparation-07ee75fd3753?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/07ee75fd3753</guid>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[interview-questions]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Tue, 17 Oct 2023 00:05:42 GMT</pubDate>
            <atom:updated>2023-12-21T04:43:02.764Z</atom:updated>
            <content:encoded><![CDATA[<p>In this Medium post, we’ll describe strategies to prepare for coding interviews if you’re a backend engineer. We won’t rehash content that has been written before (<a href="https://www.amazon.com/Cracking-Coding-Interview-Programming-Questions/dp/0984782850">for ex: here</a>), but instead offer some new pointers and tips as you prepare for a coding interview for backend software engineers.</p><h3>Preparation Stage</h3><h4>Selecting the Programming Language</h4><p>It’s important to choose a good programming language as early as possible in preparation, so you do all of your practice problems with the given language. This will be difficult if Python is NOT your strongest language (otherwise you’ll choose Python).</p><p>If you’re a backend engineer but Python is <strong>not your strongest programming language </strong>(say you’re more experienced with Java, Rust, Scala, Go, etc.), my recommendation depends on if your interview is over a a Whiteboard (in-person) or over VC with an online IDE (e.g. Coderpad or HackerRank).</p><ol><li>If the interview is <strong>over the Whiteboard</strong> (in-person), then <strong>use Python.</strong></li><li>If the interview is <strong>over VC with an online IDE</strong> (e.g. Coderpad or HackerRank, etc.) then use <strong>the language you’re most familiar</strong> with.</li></ol><p><strong>Why? Several reasons:</strong></p><ul><li>In a <strong>whiteboard setting</strong>, a major limiting factor will be how fast you can write the code out physically with your hand and marker. In that settting, the shorter syntax of Python is really advantageous.</li><li>In a <strong>whiteboard setting</strong>, you’re not expected to have the perfect syntax and whether the code compiles/runs also is less important (it may even be difficult for the interviewer to tell if the code runs/is correct). therefore your potential lack of familiarity with Python is neutralized.</li><li>In an <strong>Online / Coderpad</strong> setting, you will be expected to have the code actually running and executing. If your code has syntax errors or you are unfamilar with details of the language (say Python), it’ll be extremely glaring.</li><li>In an <strong>Online / Coderpad</strong> setting, physically writing the code out <strong>will not be a limiting factor</strong> (since you’re typing), therefore the verboseness of other languages (C++, Java, Rust, etc.) relative to Python is not a big deal. Generally the following is true: (1) typing speed &gt; (2) thinking speed &gt; (3) hand-writing speed. Therefore Python’s conciseness as a language doesn’t give you much benefit.</li></ul><p>For these reasons, I recommend backend candidates use the language they’re most familiar with if it’s online/VC/IDE interview and use Python if it’s a white boarding interview.</p><h4>General skills to practice</h4><p>Your success in a coding interview does depend on several different skills:</p><p><strong>(1) Problem Understanding/Alignment</strong>: This is your skill of understanding the problem that is being asked of you and aligning with the interviewer over expectations.</p><ul><li>How well do you ask questions to make sure you really grok the interview question and all the edge cases and nuances?</li><li>How well do you pick test cases to really test your problem understanding?</li><li>How well do you align with the interviewer on expectations (run-time, code complexity, what the solution looks like).</li></ul><p><strong>Note</strong>: one common mistake I see interviewees make is they fail to align with the interviewer on the type of solution that is expected.</p><p><strong>(2) Problem Solving + Communication</strong>: Now that you understand the problem, how effective are you in coming up with a solution both in abstract as well as on paper and communicating it to the interviewer?</p><ul><li>How well do you come up with a solution that works given the interviewer constraints/expectations?</li><li>How well do you communicate that solution to the interviewer as you’re implementing it, so they get signal on your problem-solving?</li></ul><p>(3) <strong>Implementation</strong>: You now need to get code down into the whiteboard or IDE that matches your proposed solution.</p><ul><li>How quickly and effortlessly are you able to get code down?</li><li>Do you display a mastery of the language you’ve chosen? Are you able to implement things the idiomatic way for the given language?</li></ul><p>(4) <strong>Testing + Debugging</strong>: You need to then test and debug your code. This is a very different process if it’s on the whiteboard vs. over an online IDE.</p><ul><li>How well do you come up with test cases that convince your interviewer your code is completely solid?</li><li>If there is a bug, how effectively do you find where the bug is?</li></ul><p><strong>Note</strong>: generally your debugging process will be much easier if you’ve broken down the problem into subroutines that you can test individually and have been testing each function incrementally as you develop it.</p><p>(5) <strong>Reflection on Solution (optional): </strong>Sometimes you’ll be asked to come up with O(n) for time and space and also reflect on how you’d make changes to your solution (e.g. I could make it faster if I used caching/memoization, etc.). Not every interview question will have this component, but that’s also a skill to develop.</p><h4>How to do practice interviews</h4><p>Doing practice coding interviews will be important. Try to find a friend who is also planning on interviewing and you can do practice interviews with each other. As you do the practice interviews, keep the following in mind:</p><ol><li>When practicing, don’t just focus on the outcome (did I write correct code or not) but also make sure that you exhibited all of the behaviors (e.g. aligning with the interviewer, setting up the right test cases, debugging thoughtfully).</li><li>Try to video and screen record (with your practice partner’s permission) your interview. There are several products you can use (just Google them) to record both sides’ audio as well as what’s on the screen. Watch them back and study to see if you exhibited the correct behaviors.</li><li>Practice the art of interviewer alignment. When watching back your coding interviews, check to see if you double-checked a few test cases with your interviewer to really understand the problem.</li><li>Try to isolate which skills you need the most work on and have your practice partner test those skills in particular. For ex: if you are very good at (2) problem-solving/communication but are not as good at (3) testing/debugging, doing a bunch of practice problems where you don’t debug anything won’t be as useful in helping you improve.</li></ol><h4>General Types of Problems you will See</h4><p>There are many different category of problems you might run into. Note a coding problem might involve one or multiple of these</p><p><strong>(1) Building Custom Data Structures</strong>: The interviewer will ask you to come up with some type of custom data structure to solve a problem. Make sure you review your standard LinkedLists, HashMaps, Binary Trees and how to combine and mix them together to solve various problems.</p><ul><li>Examples: Leetcode <a href="https://leetcode.com/problems/lru-cache/">146</a>, <a href="https://leetcode.com/problems/min-stack/">155</a>, <a href="https://leetcode.com/problems/time-based-key-value-store/">981</a></li></ul><p><strong>(2) Binary Search</strong>: You may have to incorporate some form of binary search into your problem solving.</p><ul><li>Examples: Leetcode <a href="https://leetcode.com/problems/missing-number/">268</a></li></ul><p>(3) <strong>String Manipulation</strong>: Get comfortable tokenizing and manipulating strings in ways to solve specific problems.</p><ul><li>Examples: Leetcode <a href="https://leetcode.com/problems/longest-substring-without-repeating-characters/">3</a>, <a href="https://leetcode.com/problems/generate-parentheses/">22</a>, <a href="https://leetcode.com/problems/simplify-path/">71</a>, <a href="https://leetcode.com/problems/restore-ip-addresses/">93</a>, <a href="https://leetcode.com/problems/word-break/">139</a></li></ul><p>(4) <strong>Graph Traversal</strong>: Get experience identifying problems as graph problems and mapping the problem as some type of traversal on the graph or doing cycle detection, etc.</p><ul><li>Examples: Leetcode <a href="https://leetcode.com/problems/evaluate-division/">399</a></li></ul><p>(5) <strong>Memoization</strong>: Memoization often fits as a “part 2” to an existing interview question. For example, you’ll be asked to solve some problem, and then apply memoization in order to make it more performant / tractable.</p><ul><li>Examples: Leetcode <a href="https://leetcode.com/problems/memoize/">2623</a></li></ul><p>(6) <strong>Using Stacks / Recursion</strong>:</p><ul><li>Examples: Leetcode <a href="https://leetcode.com/problems/flatten-binary-tree-to-linked-list/">114</a>, <a href="https://leetcode.com/problems/gas-station/">134</a>, <a href="https://leetcode.com/problems/flatten-nested-list-iterator/">341</a>, <a href="https://leetcode.com/problems/decode-string/">394</a>, <a href="https://leetcode.com/problems/evaluate-reverse-polish-notation/">150</a>, <a href="https://leetcode.com/problems/score-of-parentheses/">856</a>, <a href="https://leetcode.com/problems/design-a-stack-with-increment-operation/">1381</a></li></ul><h4>What Not to Prepare</h4><p>Knowing what not to prepare is likely just as valuable as knowing what to prepare. During your preparation, you don’t want to spend valuable time practicing problems that are highly unlikely to be asked of you during the interview.</p><p><strong>(1) Leetcode Hard Problems</strong>: These generally don’t show up because they usually don’t provide much signal to the interviewer about your ability. Leetcode Hard problems tend to focus on pattern-matching some difficult algorithm to some problem type.</p><p><strong>(2) Dynamic Programming</strong>: Generally, dynamic programming probably won’t be in your interview (or if it is, it would be the bonus solution for some question that you can solve some other way).</p><p><strong>(3) Advanced Niche Algorithms</strong>: There’s a class of optimal algorithms that you study in Computer Science class that are probably not as useful to study.</p><ul><li>For example, it’s unlikely that an interviewer will have you do a graph traversal that would require Dijkstra’s algorithm, when simple BFS/DFS/recursion would not suffice.</li><li>As another example, it’s unlikely the interviewer would ask you a problem where you’d have to do Linear Time Selection in O(n) when a much easier O(n log n) would not be acceptable.</li><li>If you ever run into one of these, and your interviewer is asking if you can do better than O(n log n ), you can always tell the interviewer, “I know there’s an O(n) solution based on Quicksort Linear Time Selection but I don’t remember the specific algorithm”.</li></ul><h3>During Interview</h3><p>Once you are in the coding interview, there are some tips to increase your overall performance. These are not necessarily directly correlated with your core coding ability but are helpful interview tips.</p><h4>Ask the interviewer how many questions</h4><p>It’s a good idea to ask the interviewer how many questions and how many parts to each question they’d ideally like to get through during the session. The interviewer will usually tell you something like, “two questions, Question #1 has 2 parts and Question #2 has 1 part” or “only one question but it has 3 parts”.</p><p>Knowing this information is helpful for you as you pace/time your interview as it’ll give you signal if you’re falling behind or you’re ahead of time.</p><p><strong>You’re ahead of time</strong>: If you’re ahead of time (for ex: the interviewer has told you there’s 2 parts to the problem and you’re already on part 2 with 30 min. to spare), then you <strong>can take less risk/variance.</strong></p><ol><li>You’ll want to break down the problem into smaller functions, test each one thoroughly before moving onto the next question, etc.</li><li>You can really thoroughly think through the design and syntax and make sure you’re writing really clean and solid code with great variable names, etc.</li></ol><p><strong>You’re behind on time</strong>: If you’re behind (for ex: there are three parts and you have ~15 min. remaining and still on part 1), then you want to <strong>increase the variance/risk</strong>. At this point you’re maybe looking to avoid a “Strong No Hire” and maybe at least salvage a “Weak No Hire” or “Weak Hire”.</p><ol><li>You’re more likely to ask for a hint to get you past whatever you’re stuck on so you can demonstrate positive signal for part 2, etc.</li><li>You can take a higher risk strategy of trying to write out the code as quickly as possible, and skimp on testing each specific subroutine in order to get a working version.</li><li>You might opt for trying a brute force solution that simply works rather than one that is more optimized.</li></ol><h4>Align with the interviewer on the constraints and a solution</h4><p>Before you even write a single line of code, you should try to align with the interviewer if your solution is reasonable. Make sure you align on which edge cases/error conditions you’ll worry about. Some examples:</p><ul><li>“I’m writing a brute force solution, is that okay with you?”</li><li>“I can implement this with memoization or without, which do you prefer?”</li><li>“Can I assume that the inputs won’t include negative numbers?”</li><li>“What happens if an entity points to itself in a cycle? Should I worry about this case?”</li></ul><p>It’s important to stay disciplined and do this exercise during the interview even if you think the interview question being asked is trivial.</p><h4>Break a large problem into smaller ones</h4><p>Imagine as part of the coding interview, you’re asked to write a function to multiply two large numbers (say 100-digit numbers) together that are represented as int[] digits (since they don’t fit as 64-bit long integers).</p><pre>private int[] multiply(int[] digits, int[] digits2);</pre><p>Many candidates will try to write the code for this as one large function. However, this approach has two problems:</p><ol><li>The code within the function will be extremely long, which will be prone to mistakes. You’ll have to iterate with nested for loops, increasing the likelihood of off-by-one error or some other issue.</li><li>It’ll be difficult for you to test different parts of the code and isolate where a potential bug/error is.</li></ol><p>Instead you can abstract away the problem. Imagine if you have two functions: (1) that adds two large numbers (represented as int[] digits) and (2) that multiplies a single digit to a large number.</p><pre>private List&lt;Integer&gt; add(List&lt;Integer&gt; digits1, List&lt;Integer&gt; digits2);<br>private List&lt;Integer&gt; multiply(List&lt;Integer&gt; digits1, int digit2);</pre><p>Assuming these two functions existed, then the implementation for multiply becomes trivially easy:</p><pre>private List&lt;Integer&gt; multiply(List&lt;Integer&gt; digits1, List&lt;Integer&gt; digits2) {<br>  List&lt;Integer&gt; result = new ArrayList&lt;Integer&gt;(<br>  <br>  for (int i = 0; i &lt; digits2.length; i++) {<br>    List&lt;Integer&gt; partialResult = multiply(digits, digits2.get(i));<br>    for (int j = 0; j &lt; i; j++) {<br>      partialResult.add(0, 0);<br>    }<br>    result = add(result, partialResult);<br>  }<br>  return result;<br>}</pre><p>From here you can then implement the subroutines add and multiply for single digits. This approach (breaking larger problem into smaller problem during the interview) offers several important advantages:</p><ol><li>You’re much <strong>less likely</strong> to make a mistake or a bug when each function you’re writing is much smaller.</li><li>If you do have a bug, it <strong>becomes much easier</strong> to test and identify where the bug is. You can test each of the subroutines by themselves to make sure they’re working 100% and conclude that a bug might exist in the parent function.</li><li><strong>Saving implementation</strong>: In some cases (if you’re interviewing on a whiteboard) your interviewer might even decide it’s not worth having you implement the subroutines. In this case, the interviewer might decide that impelmenting add or multiply with single digits is extremely trivial and move onto coding question #2, etc. This would be a huge win, as you get full credit for solving the problem, even though you only had to write a fraction of the full solution.</li></ol><h4>Hint-Asking vs. Clarification</h4><p>It’s okay to ask for hints, but the best interviewees are able to frame their “hints” as “clarification questions.” When you do this effectively, the interviewer doesn’t register you ask as “hint” but rather blames themselves for not providing a more clear question (or even gives you bonus points for checking edge cases!) There are several techniques/examples for how to do this:</p><ol><li>If you’re stuck or confused, try writing out some more test cases / examples and validating expectations.</li><li>Stall for more time by validating trivial constraints with the interviewer (e.g. if the input is 0or null).</li><li>You can often fish for a hint by simply acting confident and asking a clarifiying question. For example: “I know there’s a brute force solution, but I’m pretty sure there might be a more performant one — do you want me to search for that one, or just go with the brute force one?”</li></ol><h3>Additional Reading</h3><p>For more pro-tips on preparing for a coding interview check out <a href="https://www.amazon.com/Cracking-Coding-Interview-Programming-Questions/dp/0984782850">Cracking the Coding Interview</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=07ee75fd3753" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Mastering the Remote System Design Interview]]></title>
            <link>https://tonygwu.medium.com/mastering-the-remote-system-design-interview-de7120b9ea52?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/de7120b9ea52</guid>
            <category><![CDATA[interviews-and-insights]]></category>
            <category><![CDATA[remote-working]]></category>
            <category><![CDATA[tech]]></category>
            <category><![CDATA[system-design-interview]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Thu, 10 Dec 2020 18:31:26 GMT</pubDate>
            <atom:updated>2020-12-12T02:13:40.261Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This post is part of a series titled, “</em><a href="https://medium.com/@tonygwu/how-to-select-your-next-job-fe7eb8628156"><em>How to Select Your Next Job</em></a><em>”. I’d love to hear your feedback in the comments section below!</em></p><p>If you’re in the process of seeking a new technical job as a Software Engineer or Engineering Manager, you’ll likely have to do a System Design interview. The interview involves solving a high-level technical problem (e.g. designing some well-known product or service) and demonstrating your technical experience and intuition for coming up with a solution.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/612/1*Wyh11BnWBfBJO8P7PEzWjw.jpeg" /></figure><p>Currently, with the global pandemic and the work-from-home environment (in 2020/2021), tech companies are conducting these interviews over VC, which introduces some challenges for the interviewee. Most candidates have extensive experience interviewing in-person on a white-board, but are new to being interviewed for system design over VC. The purpose of this article is to help you best prepare and do well during these interviews.</p><h3>Why are Remote System Design Interviews Challenging?</h3><p>There are several reasons why doing a system design interview remotely over VC is challenging:</p><h4><strong>1. Unfamiliar setting and tooling</strong></h4><p>When you do a remote system design interview, you must utilize a different set of tooling (Bluejeans for VC, Google Draw or an equivalent in place of a whiteboard).</p><ul><li>Most candidates are fairly experienced at white-boarding a problem/solution to a technical problem. You can quickly draw boxes and label diagrams on a whiteboard and adjust the size easily.</li><li>While doing an interview over VC you’ll have to use a tool like LucidChart or Google Draw, that you might not be as familiar with. As a result, drawing and configuring diagrams (e.g. sizing them correctly, adjusting fonts, etc.) will slow you down.</li></ul><h4>2. Communication latency</h4><p>Communication can be challenging because of network latency on VC, as both you and the interviewer may inadvertently interrupt each other while discussion is ongoing.</p><h4><strong>3. Interviewer inexperience and lack of calibration</strong></h4><p>Most of the interviewers you’ll face will have extensive experience with interviewing candidates in person, but have more limited experience with interviewing over VC. As a result they’re likely to be less well-calibrated in evaluating remote interview performance.</p><h4><strong>4. Lack of tactile feedback from the interviewer</strong>.</h4><p>In a normal face-to-face setting, the interviewer can easily jump in with her/her own marker and work with you on the problem or provide clarification for points of confusion. This is significantly more challenging and less natural to do in a remote setting.</p><h3>Opportunity for Preparation</h3><p>However, the fact that your system design interview is being done in a remote fashion, does give you the opportunity for preparation. This document will highlight some tips to help you set up and prepare for an optimal remote interview session.</p><h4>Monitor and Sizing</h4><p><strong>Monitors</strong>. I recommend that in preparation for remote system interview, you have two monitors set up, one for the Bluejeans/Zoom video itself as well as the Components Setup scratchpad (see below), and one other monitor for the actual Google Draw / Lucidchart where you’ll conduct the system design interview. I also recommend getting a desk lamp and move the interviewer’s video overlay close to your video camera so you make eye contact when looking at them.</p><p><strong>Sizing</strong>. Play around with the size of your drawing canvas ahead of time. In a normal face-to-face interview, you have an entire large whiteboard that you are free to utilize. However, I’ve observed a lot of candidates spend precious interview time resizing their canvas to create more space as the interview progresses.</p><p>To avoid this, you should set the size of the canvas to be as big as possible such that the default font is still visible on a full-screen VC screen share. If you’re using Google Draw, I personally recommend pre-setting the page to be Custom, 30 in. by 20 in (see below).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/732/0*KeIcdCts5zgro6Ww" /><figcaption>On Google Draw, go to “File” → “Page Setup” and select “Custom”</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/914/0*3Mfd8uHVUxEOaFwO" /><figcaption>I recommend picking 30 by 20 inches on a 30&quot; monitor as the optimal size.</figcaption></figure><h4>Pre-set Components</h4><p>When using Google Draw or Lucidchart, interviewees often end up struggling to draw components quickly. Since your goal in the interview is to as quickly as possible convey conceptual concepts to the interviewer, you don’t want to waste time playing around with the tooling during the interview. In the face-to-face setting, when working with a whiteboard, you can easily draw a square/rectangle/cylinder however large you’d like and draw inter-connected arrows easily. However in a remote environment, I often see candidates spend all their time dragging and dropping boxes, resizing and adjusting colors in order to get it to show up appropriately onto the canvas.</p><p>To help mitigate this, I recommend for candidates to prepare a set of “pre-set” interview components that they can quickly copy-paste onto the interview canvas as needed, simply by using “cmd-c, cmd-v”. One of the good things about using Google Draw is that copy-pasting components works across Google Chrome browsers.</p><p>Here’s an <a href="https://docs.google.com/drawings/d/1u9M6FQO4f3iHwLYO9QAHrkB6R2an8FVTw-uLZ0Z4PWc/edit?usp=sharing">example template</a> that includes commonly used components for system design interviews (just to give you an idea). You can include:</p><ol><li>Text boxes (copy-pasteable to discuss one specific aspect of a system design)</li></ol><ul><li><strong>Math Computation</strong>. In case you want to do some math around # of machines or $$ cost required to solve a problem.</li><li><strong>Cache Definition</strong>. If you’re asked to provide what a cache looks like.</li><li><strong>Schema Definition</strong>. If you need to do some schema definitions.</li><li><strong>API Definition</strong>. In all likelihood you’ll have to define a protocol between two different components/services.</li></ul><p>2. Components (easily copy-pasteable into an interview): Databases, Services, Clients, Caches, Queues, etc.</p><p>By preparing a templated set of “components” that you can easily copy-paste into your interview window as needed it ensures: 1) You spend a minimal amount of time conveying your ideas across to the interview and 2) That you hit all of the key points you want to talk about for a given topic (ie. for schema definition, api definition, match computation, you don’t forget any of the standard things you want to cover).</p><h3>During Interview: Guiding and Collecting Feedback</h3><p>Keep in mind that the purpose of the interview is for you to <strong><em>help the interviewer collect the signal</em></strong> rather than for you to arrive at the “<strong><em>right answer</em></strong>”. This means that even if you have the perfect answer, but either your interviewer doesn’t grok/buy into it or you fail to communicate in a way they understand, then you’ll fail the interview all the same. For example, if you design the perfect system that uses all common / open source software (e.g. kafka + Redis + EC2 + Docker + Redshift, etc.) but your interviewer has worked his/her entire career at Microsoft or Facebook, and has never heard of these concepts, then it’s likely you won’t achieve success in the interview.</p><p>While shadowing interviews, I’ve observed a common failure pattern where a candidate spends most of the interview talking non-stop only to have the interviewer not get a chance to cover a major piece of signal or follow-up to the interview question they wanted to collect signal on.</p><p>However, this doesn’t mean that as an interviewee you should be passive — the interviewer is still expecting for you to drive the conversation. Let’s go through some techniques you can use to ensure that you thread this needle to success.</p><h4>Expectation-setting</h4><p>To ensure that you cover parts of the problem-space that the interviewer cares about, ensure that you level-set during the interview, but do so in a proactive way that gives him/her the impression that you have a broad set of capabilities. Here are some example level-setting statements you can make during the beginning of the interview:</p><ul><li>“This is a very interesting problem that has many different facets. Just off the top of my head, we can talk about API design, doing some computation scale analysis, data modeling for storing the Tweets, I can also talk about how to scale this or how to handle whale users. Is there a specific area you want me to cover first or dive deep into?”</li><li>“Before we start, I’m going to spend 5 minutes quickly doing some back-of-the-envelope calculation to figure out how many PetaBytes of in-memory cache and hard storage we need. Does that sound good to you?”</li><li>“After I talk about the storage layer, I can either talk about how to scale this up for billions of users or I can talk about authentication and security. I think the former is more intellectually interesting, but do you have a preference?”</li><li>“Coming from a startup background, I’m intimately familiar with the AWS stack (redshift, EC2, S3, Cloudflare). Are you okay if I proceed using terminology from that?</li></ul><p>In each of the above examples, you’re collecting signals from the interviewer about his/her general expectations, while also staying in the driver’s seat of the interview. This allows you to project an impression of being very broad and versatile, while also making sure you don’t waste much time covering any areas the interviewer is not as interested in.</p><h4>Familiarize yourself with the company’s tech</h4><p>If the company you are interviewing at has built a ton of technology in-house, it’s important that you spend some time researching what the in-house technology versions are. This is <strong><em>especially true if your interviewer has been at the given company for a long time</em></strong> (e.g. a 12-year Microsoft or Facebook veteran).</p><p>For example if you’re interviewing at the following companies, it’s probably good to familiarize yourself with some key terms (non-exhaustive):</p><ul><li><strong>Google</strong>: Spanner, BigTable, Hadoop, Firebase, Angular, TensorFlow, etc.</li><li><strong>Facebook</strong>: Presto, React/Native, Scribe, Scuba, Cassandra, HBase, HipHop, Hacklang</li><li><strong>Microsoft</strong>: Entire Azure stack, SQLServer, CosmosDB, Azure</li><li><strong>Amazon</strong>: You should know AWS like the back of your hand, RDS, Redshift</li></ul><p>Even if you don’t understand a given technology in-depth, literally just knowing the term can be helpful to quickly resolve understanding gaps during the interview. For example, if a Facebook interviewer asks you, “What’s <em>Kafka</em>?”, you can respond quickly with, “It’s basically just what you guys call <em>Scribe</em>”.</p><h4>Constant Checking-In</h4><p>In general, interviewers are trained to not reveal too much signal with respect to whether you’re doing well or not during the interview. It can be challenging to know if you’re missing something or going into sufficient depth on a specific portion of the interview (e.g. Did I cover the data model schema in enough detail? Should I write an immutable version of this API definition?).</p><p>To ensure that you’re covering the key points and giving the interviewer sufficient signal, you should constantly check-in with the interviewer to ensure that you’re on the right track. Here are some things you can say:</p><ul><li>“I’ll pause here and see if you have any questions.”</li><li>“Would you like me to write out the full API definition, or just go over it a very high-level or just move onto X?”</li><li>“This specific design does have some weaknesses however, it does heavily sacrifice latency/performance for future extensibility and cleaner abstraction. Would you like me to quantify precisely how much or propose an alternative solution that reverses this trade-off?”</li><li>“I front the database call here with a Redis LRU cache, which under the hood is implemented with a LinkedHashMap data structure. Would you like me to explain/explore the specifics of how this works under the hood?”</li></ul><h3>Pre-Interview Checklist</h3><p>To recap, here’s a checklist that you should just quickly go through before you jump into your system design interview. Set aside 15–30 minutes prior to starting your full slate of interviews to ensure that you go through each:</p><ul><li>Have two monitors set up, one is your main (interview canvas), and one is your side (put your templated components).</li><li>Do a quick copy-paste test between your “templated components” to your “main interview canvas” to make sure your clipboard is working.</li><li>Practice dialing in and sharing your screen using the given VC software the company’s provided (Bluejeans, Zoom, Webex, etc.).</li><li>Go to the bathroom if needed.</li><li>Set the interview canvas side as appropriate (30 in. by 20 in. for Google Draw).</li><li>Look up your interviewer ahead of time on LinkedIn (usually your recruiter gives you interviewer information).</li></ul><h4>References and Reading</h4><ul><li><a href="https://www.amazon.com/System-Design-Interview-insiders-Second/dp/B08CMF2CQF/">System Design Interview Insiders</a></li></ul><p>You made it to the end of the guide! Good luck with your interviews.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=de7120b9ea52" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Tips for non-Functional Interviews]]></title>
            <link>https://tonygwu.medium.com/tips-for-non-functional-interviews-e2da54a03b0e?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/e2da54a03b0e</guid>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Wed, 01 Jul 2020 23:40:46 GMT</pubDate>
            <atom:updated>2023-10-26T03:52:46.947Z</atom:updated>
            <content:encoded><![CDATA[<h3>Guide for Behavioral Interviews</h3><p><em>This post is part of a series titled, “</em><a href="https://medium.com/@tonygwu/how-to-select-your-next-job-fe7eb8628156"><em>How to Select Your Next Job</em></a><em>”. I’d love to hear your feedback in the comments section below!</em></p><p>If you’re a Software Engineer, Engineering Manager (or PM, DS) looking to go through the interview process at various tech companies, you’re likely to different types of interviews:</p><ol><li><strong>Functional Interviews</strong> test whether you have the skills for your specific job function. Examples of this are <strong>coding</strong> and<strong> system design interviews</strong> for engineers or “<strong>product sense interviews</strong>” for PMs.</li><li><strong>Behavioral Interviews</strong> test whether you have the right mindset, and experience, whether you’re a good collaborator, and will value the team and company interests.</li></ol><p>At Facebook (or any other company), there are many different sub-types of behavioral interviews, depending on who is interviewing you (e.g. your hiring manager, a XFN partner, etc.):</p><ol><li><strong>Hiring Manager Screen</strong>: This is generally an interview with the hiring manager who will likely ask you a series of questions.</li><li><strong>XFN Interview</strong>: Sometimes a PM or a Data Scientist, etc. They’ll look to understand how you collaborate.</li><li><strong>People Management</strong>: If you’re interviewing for a people management position, then it’s likely that a good bulk of your interviews will be of this form.</li><li><strong>Project Retrospectives or (“Jedis”)</strong>: these will often focus on how you’ve handled situations in the past.</li></ol><p><strong>Note</strong>: there are some interview types that focus a bit more around technical deep dives about something you’ve worked on in the past (Project Presentations). Many of the tips that we’ve discussed in thie post will also apply to those types of interviews, but <a href="https://tonygwu.medium.com/interview-prep-project-presentations-technical-retrospectives-4348a8467218">there’s a dedicated medium post to this specific type of interview</a> that you should read.</p><p>The purpose of this post is to give you a guide for how to prepare and be really successful with the Behavioral Interview.</p><h3>Question Types</h3><p>It’s important when answering behavioral questions to distinguish between different types of questions and understand what the interviewer is expecting from you in their question.</p><h4>“Tell Me When” vs. “How Do You” (or anything in between)</h4><p>Generally there are two categories of questions and answers that your interviewer will expect from you:</p><ol><li><strong>“Tell Me When”</strong> or <strong>“Give me an Example” (TMW/GAE) </strong>questions where the interviewer is expecting you to tell them a story.</li><li><strong>“How Do You” (HDY) </strong>questions where the interviewers is looking for you go give them a general framework for how to solve a common problem.</li></ol><p>Example <strong>TMW</strong> or <strong>GAE</strong> questions:</p><ul><li>Tell me when you took initiative to solve a problem.</li><li>Tell me when you had to admit to making a mistake.</li><li>Give me an example of when you failed to retain a high performer on your team.</li></ul><p>Example <strong>HDY</strong> questions:</p><ul><li>How do you manage a challenging relationship with a peer?</li><li>How would you utilize input from other functions in decision-making?</li><li>How do you create a culture of X (e.g. frugality, collaboration, etc.)</li></ul><p><strong><em>Note</em></strong><em>: there technically is a “third class” of questions that pertain to the specific role/company you’re interviewing for or about you in particular. An example of this type of question would be, “What are your weaknesses” or “Why are you interviewing at our company?”. This post won’t go into detail about answering these types of questions — I might do another post on how to handle these types of questions.</em></p><p>If the interviewer asks you a “Tell Me When” question, they’re likely looking for a <strong>story</strong>, and if they asks you a “How Do You” question, they’re likely looking for a <strong>framework</strong>.</p><h4>What Type of Question are you answering?</h4><p>Your job when answering the given question is to make sure you understand what your interviewer’s expectations are. Obviously if the interviewer phrases the questions with “Tell Me When” or “How Do You”, it gives you signal on whether they’re asking, but it’s always good to double-check. You can do so in an non-awkward organic way after the inteviewer has asked you the question.</p><ul><li>Ex: “Just to double-check, you’d like me to tell you about a time when I managed out a low performer, rather than what my general framework for handling low performers is, correct?”</li></ul><h4>Use Numbers + Rule of 3</h4><p>Since your interviewer will be frantically taking notes, it’s helpful to use numbers when talking about things.</p><p>For example, “<em>When I onboard onto a new team there are 3 things I generally try to do, #1 get to know the product and tech, #2, &lt;….&gt;, etc.</em>”</p><p>Another example, “<em>When leading this project we were trying to solve 3 main problems: #1 give our customers a better introductory experience, #2, #3, etc.</em>”</p><p>Numbering things has a lot of important benefits:</p><ol><li><strong>Helps the interviewer with note-taking/remembering</strong>. It helps the interviewer who’s taking notes keep track of what you’ve told them. If you talk quickly without notes, and an interviewer lost track of one of your bullet points, they might not realize it. However, if you’ve numbered your answers, your interviewer would realize they missed your point #2 and ask you to repeat it.</li><li><strong>You both can easily reference it in a follow-up</strong>. The interviewer can easily reference the number (e.g. “let’s double-click on #2 a little bit more” or “I’d like to elaborate a bit more on #3 if we have time…”)</li><li><strong>Helps you keep track and avoid repeating yourself</strong>. You’re more likely to keep track of what you talked about and give a more coherent and structured answer when you use numbers.</li></ol><p>In addition, I also recommend for you to keep your lists of things at 3–4 at most. If you have more than 4, that’s too many and difficult for the interviewer to understand and keep track of. If you do have more items to talk about (for ex: you have a framework for how to onboard onto a team that has 6 elements) you should bucket and group them together into logical atomic units.</p><h3>Handling TMW / GAE Questions</h3><p>If you’ve identified your question to be a TMW question, you’ll have to tell a story. You should have 8–15 stories prepared ahead of time for this moment.</p><p><strong>Please read </strong><a href="https://tonygwu.medium.com/behavioral-interviews-stories-tmw-6a61c41292c2"><strong>this post as it describes how to prepare, practice, and tell amazing stories during your interview</strong></a><strong>.</strong></p><h3>Aligning Expectations with the Interview</h3><p>Your answer to an interview question won’t really connect unless you align with the interviewer on what they expect from you.</p><h4><strong>Ensure your answer is on point</strong></h4><p>One common pitfall that many interviewees make is not answering the right question. Don’t assume that your interviewer is effective at communicating what he/she intends to ask you about. As the person being interviewed, it’s <strong><em>your job</em></strong> to make sure you take your answer in the direction the interviewer is asking for.</p><p><strong>Ask clarifying questions</strong></p><p>You can often do this by asking clarifying questions to ensure that you’re taking the question in the right direction. Examples:</p><ol><li><strong>Original Question</strong>: “How do you generally handle engineers who are low performers?” <strong>Clarifying Question</strong>: “I have a set of different approaches for senior engineers vs. more junior engineers — also it depends on the reasons for the underperformance (e.g. technical ability? lack of motivation?). I can talk about my experience with either, is there one specific type you’d like me to dig more into?”</li><li><strong>Original Question</strong>: “Can you talk about a time when you’ve handled conflict?” <strong>Clarifying Question</strong>: I’ve dealt with lots of conflict before: between engineers on the same team, between members of different teams, as well as between me and PMs. Is there a specific class/type you’d like me to elaborate on?”</li></ol><p>The primary purpose of asking clarifying questions is to ensure that you spend your time in the interview delivering signal that your interviewer is actually trying to probe. However, there’s a second important psychological thing it accomplishes— it projects to your interviewer that you have a wide breadth of experiences and can jump into. This is particularly important because your interviewer has a limited amount of time and <strong>can’t probe the entire universe of possible experiences and therefore is relying on subtle cues/heuristics from your answer to extrapolate a general impression of you</strong>. Asking clarifying questions helps trigger those heuristics to come across as well-experienced.</p><p><strong>Practice different time-length versions of answers</strong></p><p>Remember, the goal of your preparation is not to give the best possible answer, but to be in position to give your interviewer the optimal signal you can within the 30 or 45 minutes that you’re being interviewed. <strong>Your answers should be trying to squeeze as much positive signal you can per unit time</strong>. Once you’ve written down your stories/anecdotes that will be useful for a wide variety of possible questions, be prepared to give different versions of each story/answer, depending on how deep your interviewer wants to go.</p><p>Because you can’t predict what types of answers/stories your interviewer will gravitate towards and want to ask more about, it’s important that you <strong>adjust your answers based on reading your interviewer’s desires</strong>. If your interviewer wants to dig deeper on a specific part of your answer/story, be prepared to oblige him/her.</p><p>For each possible story/answer, it’s good to have a <strong>short </strong>version (30 seconds), <strong>medium</strong> (2 minutes) and <strong>long</strong> one (5–7 minutes). The short version should give a very high-level TL;DR summary of generally what happened, and one to two sentences about what the outcome was and what you learned.</p><h4>Customize your questions and answers</h4><p><strong><em>Full Reading</em></strong><em>: </em><a href="https://medium.com/@tonygwu/questions-for-your-interviewer-e64228dbe225"><em>Questions to ask your Interviewer</em></a></p><p>When it’s time for you to ask your interviewer questions, nothing is as off-putting as asking questions that could be answered easily through some basic internet research. Avoid the n00b questions like “How many employees work at Facebook?”, “What’s the tech stack like?”, “How many hours do people work?”. Demonstrate to your interviewer that you’ve put in at least a decent amount of thought in the company. If the questions you’re asking are the same questions you could ask for a different company or team, then you’re not impressing your interviewer.</p><p><strong>Base-level:</strong> <em>Customize based role</em>. For example if you’re being considered for a ML role, you can ask something like, “<em>Are ML engineers responsible for fine-tuning the performance of their models when they’ve productionized them?</em>”. This at least shows that you’re aware of the specific domain/role you’re in.</p><p><strong>Medium-level:</strong> <em>Customize based on company or team.</em> This requires you to do some research on the company you’re interviewing for. Something like this would be good if you’re interviewing at FB: “<em>I read Zuck’s public </em><a href="https://www.facebook.com/notes/mark-zuckerberg/building-global-community/10154544292806634/"><em>note</em></a><em> from 2017 about his desires to build both an informed, civically-engaged and inclusive community. Given that it’s an election year, how does Facebook trade-off the need to inform people but also have them be both engaged and inclusive sometimes results in tolerating bad actors on the platform. What framework does he lay out in order to trade these things off?</em>” The more specific and customized your question can be, the more you come off as being thoughtful about your company choice (ie. you’re not just joining FB because everyone thinks it’s cool).</p><p><strong>Advanced-level : </strong><em>Customize based on interviewer</em>. Often your recruiter will give you the names (and Linkedin links) of your interviewers ahead of time. The advanced level of customizing you can do is to do some research ahead of time on then — something like: “<em>I noticed that you’ve been at Facebook for many years and worked on Messenger Notifications up until 3 years ago. I’m actually very interested in growth and engagement, since I’m the type of person who really enjoys quantitative objectives. How does a more mature growth team like that of Messenger compare to one that is more in the early stages (like Instagram Business Growth)? As an engineer did you get to influence the messaging product? My friends always complain that notifications on Messenger were X— what were the product/business reasons that led to that decision?</em>” The more you can connect/resonate with your interviewer, the better.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e2da54a03b0e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to Evaluate a Startup (for Success)]]></title>
            <link>https://tonygwu.medium.com/how-to-evaluate-a-startup-for-success-e96c98c3a90b?source=rss-495ebce96386------2</link>
            <guid isPermaLink="false">https://medium.com/p/e96c98c3a90b</guid>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[jobs]]></category>
            <dc:creator><![CDATA[Tony Wu]]></dc:creator>
            <pubDate>Wed, 01 Jul 2020 23:40:18 GMT</pubDate>
            <atom:updated>2020-07-01T23:40:18.118Z</atom:updated>
            <content:encoded><![CDATA[<p><em>This post is part of a series titled, “</em><a href="https://medium.com/@tonygwu/how-to-select-your-next-job-fe7eb8628156"><em>How to Select Your Next Job</em></a><em>”. I’d love to hear your feedback in the comments section below!</em></p><p>If you’re a Software Engineer, Product Manager, or Data Scientist without a tech background, it can often be difficult to evaluate startups. There are hundreds of thousands out there — how do you know which one is going to succeed or fail? Which ones should you bother interviewing at? What exacerbates this challenge is that as an employee, <strong><em>you have a portfolio of one</em></strong>. As opposed to being an investor where you can spread your money among many different companies — when you’re employed, you’re exchanging your time for equity in only one startup, where you’ll likely spend at least 2–4 years working. And unless you have a financial background (i.e. you worked in investment banking, venture capital or private equity), you really aren’t likely equipped with the tools for deciding which company to pursue.</p><p>Luckily, I give laypeople an easy <em>rule of thumb</em> for evaluating the financial success of a startup they’re considering.</p><blockquote>For <strong><em>Early-Stage,</em></strong> look at <strong>Team</strong></blockquote><blockquote>For <strong><em>Mid-Stage,</em></strong> look at <strong>Market</strong></blockquote><blockquote>For <strong><em>Late-Stage,</em></strong> look at <strong>Growth</strong></blockquote><p>Note — this is not necessarily investment advice. If you’re a finance person who can do precise modeling, by all means do that. This advice is geared towards tech employees who don’t come from finance.</p><h4>Definitions</h4><p>In order for this rule of thumb to make sense, we have to first define what each of these is. I define things in the following way (don’t hate on me if you’ve heard a different definition):</p><ul><li>An <strong>early-stage</strong> startup is one that hasn’t found <a href="http://pmarchive.com/guide_to_startups_part4.html">product-market</a> fit.</li><li>A <strong>mid-stage</strong> startup is one that has achieved product-market fit, and is growing the adoption of its product within that initial market.</li><li>A <strong>late-stage</strong> startup is one that has reached saturation in its initial market, and is growing the overall size of the market and/or expanding to new markets.</li></ul><p>Note — the way that I’ve defined these terms, has nothing to do with things like funding, revenue, or headcount (per se). A 10-person startup could be late stage if it’s cornered a niche market and is expanding to other markets. A 1000-person company with billions in funding could still be early-stage if it hasn’t found product-market fit. Let’s run through a few examples:</p><ul><li><strong>Sarah</strong> and <strong>John</strong> are two entrepreneurs working on a social application to help doctors keep track of patient data. Primary care physicians and specialists can take notes on their patients and share them with each other in a HIPAA compliant way. This business requires a long sales cycle and a large army of salespeople, engineers, and compliance operations specialists. Because Sarah and John have had successful exits before and strong domain expertise, they have no problem raising money; in fact they’ve already raised up to $50M and have a team of 40 people. They’ve already built a prototype that they’ve shared with some of the doctors in their family, but to get more adoption, they’re working on a second iteration of the product. This (per my definition) is an example of an <strong>early-stage startup</strong>. Despite money raised and headcount, they haven’t found product-market fit yet.</li><li><strong>Coinbase</strong> is a startup that allows people to purchase cryptocurrencies. With 11.9 million active users who use the app to purchase Bitcoin, they are quietly taking over the reserve digital currency wallet market. This is an example of a <strong>mid-stage startup</strong>. It’s found product-market fit (a product that fits its market), and is growing its share within that market.</li><li><strong>Uber</strong> is a startup that allows users who download its app to hail a vehicle to take them somewhere. It operates in the private, for-hire, on-demand transportation market (aka the Taxi market), which prior to Uber’s existence was something like <a href="https://www.marketwatch.com/story/ride-hailing-industry-expected-to-grow-eightfold-to-285-billion-by-2030-2017-05-24">$100B</a> globally on a bookings-basis. Based on what’s been reported, the company has already occupied a <a href="https://www.google.com/url?sa=t&amp;rct=j&amp;q=&amp;esrc=s&amp;source=web&amp;cd=1&amp;cad=rja&amp;uact=8&amp;ved=0ahUKEwiz0pqfiaHXAhUB12MKHYwFApwQFggoMAA&amp;url=https%3A%2F%2Fwww.cnbc.com%2F2017%2F08%2F23%2Fuber-q2-earnings-loss-narrows-bookings-soar.html&amp;usg=AOvVaw335KvNw28gytrvRbgYb_DI">sizable</a> chunk of the market, and via disruption, is actually growing the overall size of the <a href="https://www.google.com/url?sa=t&amp;rct=j&amp;q=&amp;esrc=s&amp;source=web&amp;cd=4&amp;cad=rja&amp;uact=8&amp;ved=0ahUKEwjpidHAiaHXAhVE12MKHaYKClgQFghBMAM&amp;url=http%3A%2F%2Fwww.marketsandmarkets.com%2FPressReleases%2Fmobility-on-demand.asp&amp;usg=AOvVaw14bLn7POQ6SlefL_qUmRvn">market</a>. It’s also expanding to adjacent markets (food delivery). Based on my definition, Uber is a <strong>late-stage startup</strong>.</li></ul><p><em>At each stage, the most important factor is the likelihood the startup will get to the next stage.</em></p><h3>Early-Stage Startups: All About the Team</h3><p>As Marc Andreessen (some cool hipster VC) <a href="http://pmarchive.com/guide_to_startups_part4.html">writes</a>,</p><blockquote>The only thing that matters is getting to product/market fit. When you don’t have it, it’s all that matters. When you have it, nothing else matters.</blockquote><p>Because an early-stage startup hasn’t achieved <em>product/market-fit (PMF)</em>, when you (the engineer, pm or data scientist) are considering it, the only thing that matters is the likelihood that the company can achieve PMF. Nothing else is material in terms of evaluating the potential success of the company. Good companies at this stage are constantly shifting their product or market around, thus making an evaluation of other factors like product, market, or growth irrelevant. The only thing you should be asking yourself is “To what degree do I believe that this <strong><em>team</em></strong> can find <strong><em>product-market</em></strong> fit quickly?”. During this phase, ideas are cheap. Being dazzled by a cool product prototype is pointless — there’s no guarantee that it’d satisfy a market.</p><p>Note — when I say that team is the thing that matters, it doesn’t mean to find the <em>most pedigreed team</em>, <em>most technically competent team, or most famous team</em>. The <strong><em>quality of the team</em></strong> is determined by how quickly the team can find product-market-fit. When your initial stab at PMF doesn’t work, you have to either modify the product, or try a different market.</p><p>Consider the following example. <strong>Sarah</strong> and <strong>John</strong> are entrepreneurial superstars, with multiple exits in the past. To capture the doctor market for their startup, they’ve hired an all-star team of ex-Pharmaceutical reps who are the top ex-Merck/ex-Pfizer salespeople. They’ve also hired Staff-level ex-Google and ex-Facebook engineers to build a centralized patient information system. These engineers have contributed to open-source systems libraries like Cassandra and are capable of architecting high quality, scalable, bug-free systems that are extremely secure, which is important for housing patient data. However, because their startup hasn’t achieved product/market-fit, none of this matters. Sarah and John find that selling to doctors isn’t the right approach and that the real market should be consumers who store their own patient data on an iPhone app.</p><p>Evaluating the team at an early-stage startup, isn’t necessarily about finding genius or pedigree, or serial entrepreneurs¹. Ask yourself, “given the space this company is operating in, how confident am I in this team’s ability to find product/market fit quickly?”</p><p>The final reason <strong>team</strong> is so important at the <strong>early-stage</strong> is that the majority of the intrinsic value of the startup comes from the team. The price that someone would pay for the team in an acquihire can serve as a floor or “safety net” to what the company could fetch in an exit, provided it doesn’t achieve any form of traction.</p><h4>Questions to Ask:</h4><ul><li>Is the team well-equipped to handle change rapidly? If in finding PMF, they had to either change the product or the market, how easily could they do so? The best teams are able to change both on a dime.</li><li>How quickly can the team move? Is there some attribute about this team that gives it an unfair advantage to other pre-PMF teams trying to operate in the same space?</li><li>Does this team have some specific skill set(mobile in ’10, VR/AR in ’14, self-driving automation in ’16, etc.) that an acquirer would pay a premium for?</li></ul><h4>Questions that are Irrelevant:</h4><ul><li>What is the market opportunity for the product being built? (PMF not reached yet, so market is not defined to make a call.)</li></ul><p>[1] That’s not to say that a team of geniuses or domain experts or serial entrepreneurs is necessarily a bad thing. If the startup is in a space that requires you to be a genius to reach product/market fit quickly, then it might actually be a good thing.</p><h3>Mid-Stage Startups: All About the Market</h3><p>So you’re considering a startup that’s already found PMF. This means the company’s already shipped a minimally viable product that satisfies at least some portion of the market. What matters now is evaluating how <strong><em>good</em></strong> the market really is (and the company’s strategy for capturing it). There are many different factors to consider (here are about 10 of <a href="https://personalmba.com/10-ways-to-evaluate-a-market/">them</a>). However, if you’re an engineer like me (and not a business/finance person), it can be difficult to evaluate the startup against all of these, so I generally scope it down to just four factors:</p><ol><li><strong>Size of the market</strong></li><li><strong>Is the market growing?</strong></li><li><strong>Unit economics/metrics of the market</strong></li><li><strong>Capturability and Defensibility</strong></li></ol><h4>What is the size of the market?</h4><p>This seems like a no-brainer. Entrepreneurs or recruiters often like to brag about the size of the market their startup is disrupting (“The used car sale market is $2 Trillion!!” or “The enterprise payroll market is $400B!”).</p><p>The first point here, is that what really matters is not precisely how big the market is, but simply whether the market is <strong><em>big enough</em></strong>. Once the market is of a large enough size, diminishing returns catch up pretty fast (one company that boasts a trillion-dollar market is really not that much better than another one that boasts a hundred-billion-dollar one). The second point is that companies will define whatever addressable market they think makes them look better. It’s up to you to decide how much of that market their product is actually capable of capturing. For example, if you receive an offer from a startup that helps you mine bitcoin, they might claim their addressable market should be pegged at the entire size of all reserve currency + gold (in the trillions), and you might argue the addressable market is just that of the major cryptocurrencies bitcoin+ether (tens of billions). My general guidance on this is, for mid-stage startups, generally go with the smaller one².</p><h4>How fast is the market growing?</h4><p>This is substantially more important than overall size of the market. Choose a small-but-fast-growing market over a large-but-stagnant market every single time.</p><p>The entire mobile advertising market in 2010 was <a href="https://marketingland.com/us-digital-ad-revenues-60-billion-2015-iab-174043">$600m</a> (tiny). At the time if you were just going off of total market size, you wouldn’t have likely concluded that two startups (Facebook and Twitter) would be raking in <em>billions each </em>in mobile ad revenue <em>in a quarter</em>, within 6 years. Mobile advertising, from 2010–2015 grew more than <em>33x</em>, more than doubling each year. A rising tide lifts all ships is a truism, even with poor execution, companies performed well in this space.</p><p>Markets that are large but stagnant generally have an incumbent. For the company to grow its share, it requires competing and taking market share away from the incumbent. Markets that are naturally growing present opportunities for companies to grow without taking share away from incumbents. The former, at minimum, requires a product that is either differentiated, cheaper, or better than the incumbents’ in some facet (or more than one facet). The latter does not. <strong>For mid-stage startups with PMF, how fast the market is growing is perhaps the most important factor to consider</strong>. [2]</p><h4>What are the Unit Metrics like for the market?</h4><p>There are two specific unit metrics that are important when considering the market that your mid-stage startup is in: LTV and COA.</p><ol><li><strong><em>Lifetime Value</em></strong>: What is the lifetime value for users in said market?</li><li><strong><em>Cost of Acquisition</em></strong>: What is the cost of acquisition for customers in said market?</li></ol><p><strong>Lifetime Value</strong>:</p><ul><li>How much revenue will each customer net, on average to the business throughout the time that they are a customer of the business?</li><li>In order to compute this, you need to do some cohort analysis and understand the average churn for a customer.</li><li>You use the average revenue each customer delivers per time period divided by the average churn rate for that time period.</li><li><em>Note</em>: there are likely more advanced models for this, but for your analysis this is all you need to know.</li><li><em>For example</em>: Let’s take the market of all potential ride-sharing customers in Noe Valley in San Francisco. If the average ride-sharing customer is willing to take $50 worth of trips (in terms of gross profit), and without any promotions or marketing has a 20% chance of dropping off of a ride-sharing platform, then the average lifetime value in this market would be $250.</li><li><em>Note</em>: For any given market (and thus business) there are fixed and variable costs. You should also evaluate: 1) in all likelihood you should incorporate variable costs into your lifetime value computation and 2) to what degree even variable costs might decrease due to things like economies of scale.</li><li><em>Note</em>: Generally speaking, Lifetime value generally applies to customers within the same market, so businesses competing for the same set of customers will have the same lifetime value (so Lyft’s and Uber’s customers are likely to have the same lifetime value). This assumption breaks down if one business is able to generate some “sticky” advantage through either branding (e.g. Apple loyalists), market power (e.g. Amazon), network effects (e.g. Facebook’s social graph).</li></ul><p><strong>Cost of Acquisition</strong>:</p><ul><li>How much does it generally cost (on average) to acquire customers?</li><li>Generally for any given market there are multiple ways to acquire customers. When evaluating the startup’s market, you should try to do your diligence across a wide variety of current and possible acquisition opportunities.</li><li>This can be things like: CPA for Google/Facebook ads, Sales salaries and bonuses, Marketing expenses.</li><li>To compute this with precision, model the customer acquisition funnel and measure the cost and conversion at each point in the funnel.</li><li><em>For example</em>: The business has discovered the most cost effective way to acquire new customers is through Google Adwords. The average Cost per Click charged for an ad unit is $0.20, and once the customer lands on the website, there’s a 2% chance the user will go through the signup flow for a free trial. Finally, 20% of users who complete the free trial convert into a paying user. The average cost of acquisition for a paying customer is therefore $0.20 x 50 x 5 = $50.</li></ul><h4>What is the Capturability and Defensibility of the market?</h4><p>Given the company has found product-market fit, it’s also important to understand two key additional components of the market: “capturability” and “defensibility”.</p><ul><li>Do the dynamics of the market make it easy in particular for your product to capture (more so than other competing products)?</li><li>Once your product captures the market, how easily can the product “defend” against other products that are also trying to capture the same market?</li><li>Usually capturability and defensibility of the market go hand-in-hand. This means that in markets where your product can easily capture new customers, other products can also more easily turn around and take those customers away from you.</li><li>Think through the nature of the product-market fit that the startup’s product has delivered. Has it been achieved through easy user experience and setup? Is it because it’s priced very low? How easy would it be for a competing product to also replicate the same level of success?</li><li>Think about if the barriers to entry work in the product’s favor or against it.</li></ul><p><strong>Barriers to Entry</strong></p><p>There are many different types of barriers to entry.</p><ol><li><strong>“First-Mover” (low)</strong>. Customers prefer one product simply because it was there earlier. By being there earlier, you gain advantage and don’t incur switching costs to get your product in front of customers. For example, if you’re a content or media company, being there earlier means that you have content that others don’t have. The reason this is a weak barrier to entry is that over time it will erode unless cultivated — meaning you have to continuously spend $$ to continue to stay ahead of others.</li><li><strong>Economies of Scale (medium)</strong>. The second class of barriers to entry surround economies of scale. In this scenario, being massive gives your cost and value structure decisive advantages. Bigger = Better. Amazon, Google, Uber are all good examples of this. One common under-appreciated way in which economies of scale manifest itself is in the <strong>use of data</strong> — it’s common that the more data a product has, the better that product can end up being (for ex: Search and Newsfeed fall in this category).</li><li><strong>Network Effects (high)</strong>. This is the highest type of barrier to entry — each person on a network adds to the aggregate value of the network in a super-linear way. <em>Ex</em>: Two-sided marketplaces, Telephone, Email, Microsoft Office, Facebook.</li></ol><p>Can your mid-stage startup, that’s developed product-market fit leverage barriers of entry to its advantage? When Google was starting out, it operated in a world where the barriers to entry were low — the only ostensible advantages other technologies (physical libraries, Yahoo’s manual directory of the web) had were first-move advantages (e.g. they had built up indexing of content). Google over time turned the market into one in which economies of scale would dominate (the more searches operated on Google’s platform, the better data it had to improve its search functionality).</p><h4>You might ask…</h4><p><strong>Why shouldn’t you look at Team when evaluating startups at this </strong>Generally speaking, <strong>team</strong> does make a difference at every single stage (CEOs do after all get paid the big bucks at all stages of the startup). However, once a company has reached product-market fit, the market itself is usually what dominates the outcomes.</p><p>This point is particularly important as <a href="https://medium.com/u/fa65e64cf273">Marc Andreessen</a> and <a href="https://medium.com/u/ec9c91098b5d">Andy Rachleff</a> have <a href="https://pmarchive.com/guide_to_startups_part4.html">said</a> multiple <a href="https://www.youtube.com/watch?v=u1fPnoAF-2Y">times</a> — if you are an awesome founder in a bad market, you’re dead, but if you’re a below average founder in an amazing market, you’ll still likely build a good company.</p><p>If you’re an engineer trying to pick which startup to join, at each stage, you’re effectively trying to compute the probability that startup makes it to the next stage successfully, and once a company has found product-market fit, the overly dominating factor is the characteristics of the market.</p><p><strong>Why shouldn’t you look at Growth when evaluating Mid-Stage Startups? </strong>Good question. The easy answer is that growth can be faked at this stage. There’s no scarcity of techniques people can use to juice growth numbers. Companies can pay for app installs, Facebook and Twitter ads, offer promotions, etc. However once a startup hits a certain size it becomes unrealistic and cost prohibitive to do so.</p><h4>Questions to Ask</h4><p>If you’re an employee trying to evaluate whether to join a mid-stage startup, here are the questions to ask yourself:</p><ul><li>Given the company has product-market fit — what is the current size of the market that the existing product is a good fit for (e.g. perhaps the product works for one subset of the market but will need to be improved for the rest of the market)?</li><li>How quickly is the market growing over time?</li><li>What are the unit economics/metrics of the market? Is the Lifetime value significantly (multiples) higher than what I anticipate the cost of acquisition to be? Do people churn frequently in a way that would cause LTV to be low?</li><li>Is the product that has developed product-market fit well-positioned to make use of the barriers to entry that will develop?</li></ul><p>[2] The counterexamples to these are companies that fundamentally grow the market they operate in (e.g. Uber has grown the taxi market, Airbnb the hospitality market, i.e. people are taking Uber today who never took a taxi before). However, since we’re talking about mid-stage companies, for now I’m purely evaluating them on the likelihood they become a late-stage company and ignoring what happens once they saturate their initial market. At least for these two deca-corns, the open question of whether the market was big enough at the time is a resounding yes.</p><h3>Late Stage Startups: All about the Growth</h3><p>I define a late-stage startup as one that has already reached a significant share of its natural market, and is now looking to either: 1) start to actually grow the natural size of its intrinsic natural market or 2) fundamentally play in additional markets. Where to draw the cut-off line for when a startup pivots from a mid-stage to a late-stage startup is often blurry/unclear. Here are some examples for delineation:</p><ul><li>For Yahoo, they pivoted to becoming late-stage around 1996/1997, when the company moved their original goal of organizing the internet around hierarchy, and started investing in the overall broader growth of the internet (e.g. acquiring geocities, RocketMail, and building Yahoo! Groups and Messenger).</li><li>For Uber, the company started morphing into a late-stage startup around 2014/2015, with the launch and investment of UberFresh (later Eats), investments in Self-driving, and expansion of UberRush.</li><li>For Google, I use 2004/2005 as the delineation, as this coincided with several key moves (development of Gmail, acquisition of Youtube) that signified the business was expanding beyond its core market.</li></ul><p>So why do I encourage employees looking to join a late-stage startup to focus primarily on growth (rather than market or team)? A few reasons:</p><ol><li>Startups that reach late-stage have already demonstrated mastery over their given market, and have had a team in-place that knew how to grow a product to market.</li><li>Many of the subsequent expansions of the business beyond the initial market require well-capitalization and success within the given market that is measured through its growth.</li><li>Once the company reaches late-stage, then at least a sizable amount of success of the company is already guaranteed. The employee’s prospects therefore are primarily tied to how quickly that company is growing.</li><li>Things like “team” and “market” become less meaningful, as the company has a very large “team”, and a diversified set of markets it is pursuing. Therefore evaluating them has less of an outsized impact on your success at the company.</li><li>Late-stage companies generally can no longer “fake” growth through unscalable means (paying for the growth). What you’re looking for is prolonged, sustainable, proven growth at large scales. Early-stage or Mid-stage companies can often see growth spurts only to drop out (think Yo or Yik Yak). But such growth surges that occur at late-stage are more likely to be legitimate.</li></ol><p>So what are some ways to evaluate whether the late-stage company you’re joining is growing quickly or slowly?</p><ul><li>LinkedIn Premium will tell you if the late-stage company is hiring aggressively and growing Headcount at a fast pace.</li><li>Google Trends will tell you if there is very large growth in organic search interest.</li><li>Premium analysis tools like AppAnnie and Evercore ISI will tell you download, usage, application metrics.</li></ul><p>There you have it! You don’t have to be a valuation expert, have an MBA, or be ex-Finance to evaluate the startup. Just follow these three rules: For <strong>early-stage startups</strong> evaluate <strong><em>team</em></strong>, for <strong>mid-stage startups</strong> evaluate <strong><em>market</em></strong>, for <strong>late-stage startups</strong> evaluate <strong><em>growth</em></strong>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e96c98c3a90b" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>