<?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[The CTO’s toolbox - Medium]]></title>
        <description><![CDATA[What are the concerns of a technical leader? What should you be doing and avoiding? There is no script for this and every job is unique. This blog is an accumulation of my experiences and advice - Medium]]></description>
        <link>https://medium.com/a-technical-leaders-toolbox?source=rss----fe5f447ffd93---4</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>The CTO’s toolbox - Medium</title>
            <link>https://medium.com/a-technical-leaders-toolbox?source=rss----fe5f447ffd93---4</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 14:20:58 GMT</lastBuildDate>
        <atom:link href="https://medium.com/feed/a-technical-leaders-toolbox" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Quality: an often misunderstood concept]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/quality-an-often-misunderstood-concept-3acf2b31a17d?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/3acf2b31a17d</guid>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[quality-assurance]]></category>
            <category><![CDATA[quality]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[engineering-management]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Mon, 20 Feb 2023 19:59:03 GMT</pubDate>
            <atom:updated>2023-02-20T19:59:45.572Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mHPdNEhwqK8ZTIx1YcTw3w.jpeg" /></figure><p>Many years ago, I wrote a tongue in cheek article that ‘<a href="https://siddharth-ram.medium.com/mediocrity-is-underrated-or-better-is-the-enemy-of-good-enough-b3ac52642344">mediocrity is underrated</a>’. (it was written around 2011). It got posted on hacker news, and you can read that discussion <a href="https://news.ycombinator.com/item?id=827796">here</a>. It will read as being very cynical, and it sure is. But there is one part of it the article that deserves some thought</p><p><strong>💡 <em>You often want to be as good as absolutely necessary, but no better.</em></strong></p><p>I expect most of us might not either agree or understand the intent of this statement. To explain what I mean, we need to understand what quality is. The pioneer in Quality was Dr. <a href="https://en.wikipedia.org/wiki/W._Edwards_Deming">W. Edwards Deming</a>. He pioneered approaches to quality control, and is revered particularly in Japan — his teachings resulted in the rise of high quality cars made by Japanese car manufacturers.</p><p>My two word takeaway from Dr. Deming on what quality is:</p><h3>📖 <strong>Meets Specifications</strong></h3><p>That’s it. If you have met specs, you have delivered quality. You know why Toyota took such a lead in the 80s? They met specs better than anyone else. That’s it. Customers would refuse to take delivery of cars/parts not manufactured in Japan because ‘meeting specs’ with smaller variance meant that the cars or parts would just be more reliable, for a longer period of time.</p><p>Expunge words like ‘perfection’ from your vocabulary. Perfection has nothing to do with being perfect (Seth Godin). Perfect means meeting spec.</p><p>Of course, this means that the specs need to be defined. If there are no specs defined, then the question of ‘have we delivered quality’ is meaningless. There is no framework to use. Almost with certainty, not having specs means that you would not have met expectations because they have not been set. So if you find yourselves in a place where there are no quality specs written and agreed to, the question of quality does not arise. We deliver what we deliver. Which is unlikely to meet expectations, because there were no specs instantiating the expectations.</p><p>So what are the specifications for Software? They are generally covered in our non-functional requirements: see my writeup on Operational and Engineering excellence <a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-6bc494d484cc">here</a>, but in a nutshell</p><p><strong>Operational Excellence</strong> is defined as the work done before and after code is deployed into production to maintain or improve operational quality.</p><p><strong>Engineering Excellence</strong> is defined as the work required to create, deploy and maintain high quality products into production with speed and agility.</p><p>The goal of the specification is to help us <em>improve continuously</em> and demonstrate that we are getting closer to these outcomes.</p><h3>Should you be beating Specs?</h3><p>Once you have met specifications, going beyond that is of questionable value. Let’s say that our code coverage criteria is 70%. What is the incremental value of going to 80%? To 90%?</p><p>In some cases (e.g. the most often called module/function in the system or the place where we cannot afford to screw up) it is worth it. But for the rest: increasing this number will be a lot of additional work for relatively low value. Would we not have been better off moving to next project to unlock more customer value?</p><h3>🔥 <strong>Takeway</strong>: Beat specs ~10% of the time. Meet specs 90% of the time</h3><h3>It is specs all the way down</h3><p>The software we deliver is based on specification we get, typically from the product team. So how do we measure the quality of the specifications produced by the product team?</p><p>Engineering often has a binary notions of ‘correct’ and ‘incorrect’ tied to it. On the product side, it is way more squishy. We talk with customers &amp; partner teams, understand behaviors and expectations and then write a spec (typically a Product Requirements Document, a PRD). How can we measure the quality of the PRD? It is not easy, and we need to agree on it, iterate on it, put it in front of customers and then test whether we built what the expected. This is why we need A/B testing and rapid iteration as we understand better. How do we know that what we built is ‘easy’ for the customers to use? How do we have a specification for ‘intuitive’?</p><p>In general, adjectives like ‘easy’ are inappropriate for putting in a spec. There is no way to measure ‘easy’ or ‘fast’ or ‘simple’ — there have to be specific metrics associated with it.</p><h3>Company specifications</h3><p>Most companies have OKR’s — an annual goal that documents the planned outcomes at the end of the year.</p><p>Of course, these specs can be somewhat unstable. The Business climate changes. Competition changes. Customer expectations change. The specs need to address these changes. The CEO at my last company used to quote something attributed to General Dwight D. Eisenhower</p><h3>🔥 Plans are nothing. Planning is everything</h3><p>stated otherwise, ‘No plans survive contact with ground reality’. But it is important to plan anyway and have a north star.</p><h3>In Conclusion</h3><p>Quality is about meeting specs. Choose to beat them infrequently. Choose to meet them always.</p><p>Engineering specs can be quite precise. It gets harder to be precise as you move away from a rigorous practice of engineering to more of the art of product and predicting company outcomes</p><p>Practice continuous improvement — and measure the improvement. The goal will take care of itself if we are practicing and demonstrating continuous improvement.</p><p><a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-6bc494d484cc">Table of Contents</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3acf2b31a17d" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/quality-an-often-misunderstood-concept-3acf2b31a17d">Quality: an often misunderstood concept</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Interviewing for leadership positions]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/interviewing-for-leadership-positions-13fc2f4778b5?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/13fc2f4778b5</guid>
            <category><![CDATA[engineering-culture]]></category>
            <category><![CDATA[technical-interview]]></category>
            <category><![CDATA[engineering-mangement]]></category>
            <category><![CDATA[cto]]></category>
            <category><![CDATA[interview-tips]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Wed, 16 Feb 2022 00:30:47 GMT</pubDate>
            <atom:updated>2022-05-17T18:39:46.074Z</atom:updated>
            <content:encoded><![CDATA[<p>In <a href="https://medium.com/a-technical-leaders-toolbox/how-to-hire-b604f4f9d331">how to hire</a> I shared the practices I use for hiring. My preference is to ask candidates to work on a take home problem, and then base the interview on how they approached solving the problem. It is similar for leadership positions as well.</p><p>When the interviews are for more senior positions the take home problem becomes more ambiguous. It is not about solving a particular technical problem: at leadership levels, it is overweighted on your true north, your leadership style and where your energies are. There is no black and white ‘you did it in O(n²)‘ kind of truthiness. There is a lot more nuance.</p><p>Here is an actual case study that I ask Director of Engineering candidates (or higher) to work on:</p><blockquote>Case Study</blockquote><blockquote><strong>About this document:<br></strong>This document represents a take home assignment that will form the basis of this interview. The goal is to make sure you have time to think about the problem statement and share out your approach to solving the problem</blockquote><blockquote><strong>Some guidelines for presenting:<br></strong>1.Please start by introducing yourself — we will include a number of people who don’t know who you are before the interview<br>2.Please make any assumptions necessary to build out the scenario. The scenario is deliberately open ended. State the assumptions you have made. <br>3.The presentation will constitute the basis for deep dives with the team. The Craft demo is your time — only time we will interrupt you is to ask clarifying questions</blockquote><blockquote><strong>The scenario<br>Abacus Inc.</strong> is a company that specializes in learning and development in young children (3–8). <br>The company is a decade old. They have two main product lines: <br>1) A line of internet connected toys that get new behaviour downloaded to them <br>2) Online educational software that helps children learn by playing online games.<br>You have been hired to lead application development covering both product lines. The CEO has the following concerns that she’d like you to address</blockquote><blockquote>1.<strong>Remote work</strong>. The team is distributed across multiple geographies. She feels that there is inadequate communication and difficulty maintaining Abacus’ culture across multiple geos (Abacus has developers in China, Chile and the UK)<br>2. <strong>Productivity</strong>. She wants to understand productivity and how it can be improved<br>3. <strong>Pace</strong>. She is unhappy with the pace of development. Even though the company is cloud based, she feels that the best companies in the world move much faster than Abacus does — in fact, Abacus is falling behind its competitors on the speed at which releases take place. She wonders if we need to further modernize technology. The CTO pops in <br>at this time and shares that ..’Abacus uses state of the art, scalable EC2 instances and we even deploy every two weeks.’ He disagrees that pace is a problem.<br>4.<strong>The Analytics pipeline</strong>. Currently, there are separate leaders for the toys line and the online games with separate teams (including separate business development and product management teams). Data is stored in multiple databases (Postgres, SQL Server, MongoDB, TimeSeries DB). They are moved into a data warehouse using standard ETL stored procedures. Unfortunately, data moves only once every 24 hours, disallowing ‘real time’ decisioning. Even worse, the visualization tool requires specialized knowledge in order to build dashboards. The CEO wants self service dashboards<br>5.<strong>API’s</strong>. They are fragmented and inconsistent. Though there is a written down policy on API’s, there is little governance. Consequently, customers/partners who integrate via API’s are a discontent lot<br>6.<strong>Front End development.</strong> Much of the front end development is server side. The CEO has heard that Client side development is all the rage and wants to move to it.</blockquote><blockquote>Her directions to you are ‘Figure out all these problems. Come back with a plan. Let me know what is needed to implement the plans. Since I am an Engineer, I would like to see the technical details as well.’</blockquote><p>The case study is deliberately very open ended, covering everything from culture, technology and process. There is really no right and wrong in what we are looking for — we are looking for how the candidate approaches problem solving, their ability to present a cogent case for the problem space and its solution. This leads to a detailed discussion of the pros and cons.</p><p>Several things surprise me about these interviews, where many candidates fail to do well.</p><p>The first mistake is not spending time thinking about the problem. You have to really understand the problem space to build out a solution.</p><p>The second mistake is leading with where you energies are at. If you are a deep technologist and you ignore other parts of the problem altogether, you are going to miss the mark</p><p>The final common problem is not taking the time to actually read the instructions. For example, the very first line is ‘start by introducing yourself’ — and it astonishes me how many choose to bypass this altogether, losing out on a chance to create a real connection with the audience.</p><p>How would you go about presenting a solution to this case study? Send me a response and you could be in line for an interview at Inflection!</p><p>←<a href="https://medium.com/a-technical-leaders-toolbox/how-to-hire-b604f4f9d331">How to Hire</a>, ↑<a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-table-of-contents-6bc494d484cc">Table Of Contents</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=13fc2f4778b5" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/interviewing-for-leadership-positions-13fc2f4778b5">Interviewing for leadership positions</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Working efficiently in a remote work world]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/working-efficiently-in-a-remote-work-world-5be8ec6cff9f?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/5be8ec6cff9f</guid>
            <category><![CDATA[engineering-mangement]]></category>
            <category><![CDATA[cto]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[remote-working]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Fri, 04 Feb 2022 21:56:50 GMT</pubDate>
            <atom:updated>2022-11-07T18:41:38.156Z</atom:updated>
            <content:encoded><![CDATA[<p>Inflection moved to being a fully remote company very early into the Pandemic (March 2020). We were already heading in this direction and the pandemic accelerated it. Our decision for remote work predated the pandemic and was based on a simple plan:</p><p><strong>We go to where the brightest engineers are, rather than looking for talent in the locations where we have offices.</strong></p><p>Once this was agreed to by the leadership team, we started making changes in everything from pay grades, benefits, tooling — and we had to create a new policy for how we define our working hours. This is a place where I see companies stumble a bit.</p><p>I like to start with principles, and then flesh out a plan based on the principles. Here are the principles that inflection.com uses:</p><ul><li>We prefer working asynchronously. Asynchronous working hours allow us to work when we want and is more efficient.</li><li>Synchronous hours are reserved for times when we get significant value out of our meeting times</li></ul><p>It is not that traditional (synchronous) meetings are a bad thing — they are just overused. So we have defined a matrix of the synchronous meeting hours and when they should be used.</p><p>Attracting the best talent means giving them the maximum flexibility, while ensuring that we are doing right for our customers and our business. Large companies fall into traps of having large meetings that serve little purpose. Right sizing synchronous time ensure that we are solving for both business outcomes and for employees.</p><p>The following guidelines and direction fell out of these principles:</p><h4>Right sizing meetings</h4><h4>Synchronous hours are good for</h4><p>(a) Increasing our shared consciousness. All hands meetings (monthly) give updates to the entire company on our customer, shareholder and employee outcomes and directions</p><p>(b) Standups, Scrums, retrospectives, quarterly planning — these are what needs broad participation</p><p>(c ) Time put aside to have fun. For example, we have a virtual happy hour once a month (local time zones) where we play board games</p><h4>Collaboration Hours</h4><p>Different time zones cause problems with standardizing work hours. Inflection has the following policies in place for collaboration hours</p><ol><li><strong>All Employees, irrespective of location, share common work hours of 8 AM PST to 10 AM PST</strong>. It is expected that all employees will be able to work during these hours. This helps ensure that hours overlap for collaboration. This means Brazil has 1 pm-3 pm for meeting hours and those in Ukraine have to work 6 pm — 8 pm (EET) .</li></ol><p>So if we hire in a different geo, we ask the potential employee ‘ Will your life allow for working 8–10AM Pacific Standard Time’? If the answer is no, then they are not a good fit for us.</p><p><strong>2. Cross site collaboration outside these hours is discouraged</strong>. Teams can have collaboration hours locally outside those hours, but we discourage meetings that cross geographies outside these hours (unless they are in compatible time zones). This is to ensure that employees and contractors have the right work-life balance.</p><p>Outside these working hours, we have no policy on when employees should work. That is a decision made in collaboration with the local manager, who sets policies with employees in that geo. So if someone wants to train for a marathon and is taking noon-2 pm off daily, that is their call. All we care about is that the work gets done. Employees pick the hours they want to work. Everyone can see key metrics using jellyfish.co, but that will be a post for another time.</p><p>One of our directors in Ukraine is a night owl. He attends some of our meetings that are at 3 AM local time. He starts working around 6 PM EET. That is completely his call, and is consistent with our policies.</p><p>Allowing greater independence in work hours and providing a framework for remote work has raised our employee engagement significantly, and allows them to juggle home and work life, guilt free.</p><p><em>↑</em><a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-table-of-contents-6bc494d484cc"><em>Table Of Contents</em></a><em> |←</em><a href="https://medium.com/a-technical-leaders-toolbox/engineering-cheat-days-how-to-fire-up-creativity-6d3e7dde33bd"><em>Engineering cheat days</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5be8ec6cff9f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/working-efficiently-in-a-remote-work-world-5be8ec6cff9f">Working efficiently in a remote work world</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Engineering Cheat days: How to fire up creativity]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/engineering-cheat-days-how-to-fire-up-creativity-6d3e7dde33bd?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/6d3e7dde33bd</guid>
            <category><![CDATA[engineering-management]]></category>
            <category><![CDATA[hackathons]]></category>
            <category><![CDATA[cto]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Sun, 21 Nov 2021 21:22:29 GMT</pubDate>
            <atom:updated>2021-11-21T21:29:05.615Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*E_TzSbyLjC9M72aPVIE9Mg.jpeg" /></figure><p>I have found an effective way for managing what I eat. I restrict what I eat and when I eat 6 days of the week. But those rules are off for 24 hours, my ‘cheat day’— Friday evening to Saturday evening I eat what I want to eat, when I want to eat it. This allows me to write down all the cravings I have had during the week, and I clear that ‘backlog’ on my cheat day.</p><p>We follow a similar philosophy at Inflection.com. We have quarterly planning, epics, backlogs and stories to work on. These are the equivalent of 6 days of restricted eating. Along the way, we write down all the things that we need to fix, do better and experiment with. And once a quarter, we have our ‘cheat days’ — 4 days of throwing out all structure. No backlogs, no managers or PM’s breathing down your neck. Engineers do what they would like to do, and we make it a fun event as well(but within some guidelines).</p><p>I had helped start a similar program at Intuit, and we have the following rules at Inflection for the event we call ShipIt!</p><ul><li><strong>A bottoms up event</strong>. Managers, PM’s , execs have no say in what engineers work on during ShipIt. They offer ideas, but engineers decide what to work on</li><li><strong>They are centered around a theme</strong>, or sometimes a ‘grand challenge’ — anything that pertains to that them or the grand challenge can be worked on</li><li><strong>The ‘Fedex model’ applies</strong> — You have to deliver the changes in the week, preferably launched dark in production. (This used to be a requirement till I realized that it was preventing engineers from taking on more challenging tasks. This also means that this time cannot be used as a ‘learning’ time — you have to deliver your work to production</li><li><strong>Participation in mandatory</strong> — unless an engineer is <a href="https://medium.com/a-technical-leaders-toolbox/establish-an-on-call-program-2fb5a1d666c8">on call</a> or there is a production priority incident. Engineers have to ask for permission to be excluded</li></ul><h4>Structuring the cheat days</h4><ul><li>Distraction free — Execs, Managers and PM’s know that they are not allowed to disturb engineers during ShipIt. Getting buy in from all the execs on this is key!</li><li>Food is taken care of — Engineers get food delivery coupons to use so they can order food from their favorite restaurant.</li><li>Work with someone you don’t know- I encourage engineers to collaborate with someone they have not worked with previously</li><li>Fun is encouraged — everything from posting photos of food, to holiday themes in the winter Shipit!</li></ul><h4>The conclusion — Lightning talks</h4><p>Each team gets 5 minutes to share what their work on the last day of the event. All the executives and PM’s attend the shareout. The five minute constraint is deliberate — it teaches engineers to deliver their key message in a short time. Most shareouts have a video recording of a demo to save on time.</p><p>I find the outcome of teams is amazing. They deliver more in a compressed amount of time than any other time. And this includes taking on the most complex problems in the codebase — the things that everyone tries to avoid otherwise. One team made a significant breakthrough in observability during this time. Other teams found a way to improve performance of certain common workflows 3x. Significant, roadmap changing work has come out of this time.</p><p>Each team that moves their work into production gets a bonus. Depending on the work done and the value, planned work gets readjusted so that teams may deliver this previously unplanned activity.</p><p>As a CTO, your job includes bringing out the best in your team. A defined program to break away from business as usual helps motivate the team and clear debt.</p><p><em>↑</em><a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-table-of-contents-6bc494d484cc"><em>Table Of Contents</em></a><em> |←</em><a href="https://medium.com/a-technical-leaders-toolbox/beware-of-proxies-4fc3a8a87b4d"><em>Beware of proxies</em></a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6d3e7dde33bd" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/engineering-cheat-days-how-to-fire-up-creativity-6d3e7dde33bd">Engineering Cheat days: How to fire up creativity</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Establish an on-call program]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/establish-an-on-call-program-2fb5a1d666c8?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/2fb5a1d666c8</guid>
            <category><![CDATA[incident-management]]></category>
            <category><![CDATA[engineering-mangement]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Fri, 17 Sep 2021 00:39:06 GMT</pubDate>
            <atom:updated>2021-09-17T00:46:43.247Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vd201Jmc4cuM5W5OC_ULgw.jpeg" /></figure><p>I like having an on call program. I participate in it. Not all engineers like it.</p><p>Here are the reasons I think an on call program is important</p><h4>1. It brings engineers closer to customers</h4><p>If customer centricity is important to you — and it should be, you need to find ways of bringing engineers closer to customers. There are layers of proxies created in organizations (for good reasons) — but have the shadow of increasing the distance between customers and engineers, as I have covered in <a href="https://medium.com/a-technical-leaders-toolbox/beware-of-proxies-4fc3a8a87b4d">beware of proxies</a>. An on call program helps bridge the gap by having engineers work directly with customers and customer support experts.</p><h4><strong>2. It forces us to address deficiencies in the system</strong></h4><p>In many organizations, Customer support software is the unglamorous area. Tools are decrepit, and are not paid attention to. A rotation through customer support has always been eye opening for me. Who better than engineers to work on improving the software in the back office?</p><p>It also forces us to address monitoring, troubleshooting and observability of the system. If engineers dread the idea of going on call, you, as the CTO, are not investing sufficiently in this area.</p><h4><strong>3. It broadens the horizons of engineers</strong></h4><p>Knowledge tends to get silo’ed. By having engineers on call, they are forced to learn about areas that they otherwise would not have, and talk with other engineers that they would not normally. This is helpful.</p><p>At Inflection, all engineers go through on call. Not all engineers will look forward to being on call, and that is understandable. It is uncomfortable. You are unsure how to trace some problems. Your tooling might not be very good. If you are the CTO (and your engineering team is ≤ 100) you <strong>absolutely </strong>need to participate in on call. Nothing like some hands on experience with the state of tooling/monitoring to change your priorities. I set up our Operational Excellence team after an eye opening rotation through on call.</p><p>Some companies offer bonuses for being on call. (we do not). I’d like our on call program to get to a point where engineers look forward to being on call.</p><h4><strong>The ‘Follow the sun’ model can be very helpful</strong></h4><p>We have operations in Brazil (PST +5) and Ukraine (PST +10). This allows us to do seamless handovers from one location to another, and not have on call (except in rare circumstances) past regular working hours</p><p>→ <a href="https://siddharth-ram.medium.com/the-developer-nps-why-are-you-not-measuring-this-cdf942703aa8">Why are you not measuring Developer NPS?</a>, ↑<a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-table-of-contents-6bc494d484cc">Table Of Contents</a>,</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2fb5a1d666c8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/establish-an-on-call-program-2fb5a1d666c8">Establish an on-call program</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why I prefer Serverless]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/why-i-prefer-serverless-c7d865484633?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/c7d865484633</guid>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[serverless-architecture]]></category>
            <category><![CDATA[aws-lambda]]></category>
            <category><![CDATA[cto]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Thu, 26 Aug 2021 17:42:09 GMT</pubDate>
            <atom:updated>2022-06-08T16:50:23.616Z</atom:updated>
            <content:encoded><![CDATA[<p>I have worked and built a significant amount using K8s. I was lucky to have worked with some of the best k8s technologists you can find anywhere during my time at Intuit. Then I moved to a startup. As we do, we take the stories and experiences with us when we move, and replay stories from the same book.</p><p>This is not hate on k8s — But it is not always the right answer. Just like Serverless is not always the right answer. In decisions like these, context is king. An important part of taking on a new job is figuring out what should be unlearnt. And unlearning k8s was one of the things that I had to do. What works at large scales does not necessarily reproduce well at smaller scales.</p><p>So what was the problem? At larger companies there is typically a ‘platform’ team that is responsible for developer productivity — building tooling that enables other engineers to move fast. At the smaller scale of a startup, there is no such team. we can move faster by thinking of our cloud vendor — AWS - as our platform. So the question is: what will allow a team to move the fastest given this constraint?</p><p>Let’s start with this picture from the Cloud Native Computing Foundation for the k8s landscape:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*tUnl3VJ_QeuL46KUNR1lrA.png" /><figcaption>(Note: The image got truncated because it is too large. <a href="https://landscape.cncf.io/">https://landscape.cncf.io/</a>)</figcaption></figure><p>I have 3 letters for this diagram. W,T, F. Who has time to draw up these diagrams, and what decision tree does one have to follow in making choices from this picture? This is nuts. Larger companies can afford the time and resources to have lots of meetings to discuss each box. I just need to deliver value.</p><p>k8s was already in use when I joined the startup. The complexity of this diagram showed up in small and large ways in our k8s clusters. Deployment of our k8s clusters turned out to be an event, with 4–5 engineers starting at pixels. (to be fair, some of it was reflective of the specific implementation — but again, the diagram above does not help). Troubleshooting was problematic — we had only a couple of engineers who could really dive into details and understand what was going on when a cluster was sick. Above all, there was little interest in the team-or from me — in spending the time to become k8s experts. We are not in the container management business. We are in the deliver value to customers via products business.</p><p>We had a choice: Go hire some engineers with k8s expertise. Or dump it, and change course to Serverless. Even if we chose to hire k8s experts, we would still deploy some software using Serverless. If that is the case, why bother with k8s at all?</p><p>It took some courage, but we decide to dump k8s and move all software to Lambda. (We always use our<a href="https://siddharth-ram.medium.com/architecture-decision-logging-5a1399115c83"> Architecture decisioning process</a> to make such decisions of importance) We built out a serverless paved road (soon to be open sourced) with tooling that made it easy for engineers to deploy their first serverless feature in less than half a day. The <a href="https://medium.com/a-technical-leaders-toolbox/the-developer-nps-why-are-you-not-measuring-this-cdf942703aa8">developer NPS</a> improved significantly for teams that adopted it (there were some teams that could not because monolith). Troubleshooting improved dramatically, aided by observability implementation using DataDog. And most importantly, we moved from releases being a milestone to a <strong>complete non-event — </strong>When something was ready to be released, it was released. One incident stood out to me: We had a lambda service which had a bug in production. The fix was released in about 10 minutes. Engineers on serverless release on demand.</p><p>I do not claim that the same cannot be done with k8s. It is just less efficient for a smaller team. The more that can be outsourced to Amazon, the more efficient we are. Embracing the constraints imposed by operating at smaller scales opens new doors. In our case, standing on the shoulders of the AWS behemoth opened new doors to accelerating velocity.</p><p>One of the core architecture principles at Inflection.com is to move from IaaS to Managed Services (PaaS). Our data pipeline, compute and storage all leverage PaaS as much as possible. It gives us a leg up on competition which is slow to adopt PaaS.</p><h4><strong>Reasons given for avoiding serverless</strong></h4><h4>‘It causes vendor lock in’</h4><p>This is the easiest for me to dismiss. I spend close to zero time thinking about this. AWS is our platform. Worrying about vendor lock-in is a waste of time. <a href="https://medium.com/a-technical-leaders-toolbox/beware-of-abstraction-layers-9047190716d4">Are you going to build facades</a> and avoid using every managed service on your cloud platform? Good Luck. All you will do is slow down your team. I choose to move fast by leveraging our cloud vendor, and do not lose a wink of sleep worrying about lock in. I am happy paying AWS to manage all the infrastructure, so we can focus on our domain logic.</p><h4>‘Serverless is too expensive’</h4><p>Definitely something in this. <a href="https://aws.amazon.com/lambda/pricing/">The cost for lambda</a> is based on number of invocations and the duration for every GB-Second. There is definitely a point wherein if you exceed the number of invocations + Time*GB-Seconds, you will be better off renting a CPU and not pay per invocation. Trek10 has a good writeup on the pros and cons of <a href="https://www.trek10.com/blog/fargate-vs-lambda">Lambda vs Fargate</a>, and generally shows better pricing for Fargate as requests/sec increase. But the picture is actually much more complicated. You have to take the cost of the API gateway into account for Lambda. Adding capacity in Fargate is a step function, not a smooth curve like Lambda. And then there is provisioned capacity in Lambda — if you are using this, it needs to be factored in. You do not pay for idle applications in Lambda unlike Fargate — how do you factor that in?</p><p>I like to think about these discussion in the <a href="https://medium.com/a-technical-leaders-toolbox/the-value-to-cost-ratio-ed452d36248d">Value to Cost Ratio</a> terms. How can the team move the fastest and deliver value to customers, and how do you factor that into technology decisions as opposed to just worrying about the bottom line? IMO, Serverless delivers amazing Value to Cost Ratio in a way k8s did not for us. If you can make a compelling argument for k8s in these terms, by all means, use k8s.</p><h4>“You will not be able to build all your software on Serverless”</h4><p>This one might be true for some teams. This is driven by two “limitations” (as of Sept 2021)</p><p>(1) Max memory allocation in Lambda is 10GB — if you need more, you are out of luck and</p><p>(2) A 15 minute timeout. If you have long running jobs, Lambda is not a good solution for you.</p><p>I have not hit either limit thus far, but may in in the future. And that is ok. If I can do 99% of my work the most efficient way I know how, I am ok doing the other 1% in a different, less optimized way (by less optimized I mean not spending much time on customization/automation for that pipeline)</p><p><strong>Fat and Thin Lambdas</strong></p><p>Once we decided to try out Serverless, a debate raged about ‘Thin’ vs ‘Fat’ Lambdas. Should we consider Lambdas as a ‘Function as a service’ or as a decomposition unit for microservices? There is no right or wrong answer here. <a href="https://www.linkedin.com/in/jeff-blasius-4b19741a/">Jeff</a>, my colleague, prepared this analysis after research and consultations with our friends at AWS:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*52F8hKiBxnBqR23JRtHigg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FvnWYgzRNnLouas7NyFczw.png" /><figcaption>Analysis of Fat vs thin Lambdas ( Inflection.com 2021)</figcaption></figure><p>We use both modes: For decomposition, the fat lambda pattern is used. For new microservices, a more FaaS model is generally adopted. (No, we do not call it a FaaS-ist model)</p><h3>The Toolchain</h3><ul><li><a href="http://datadoghq.com/">DataDog</a> (Observability, performance monitoring, tracing, logging)</li><li><a href="http://sedai.io/">Sedai.io </a>— gives us automated tuning of serverless functions (Note: I am an advisor to Sedai)</li><li><a href="http://serverless.com/">Serverless</a>.com &amp; <a href="https://github.com/features/actions">Github Actions</a> for deployments</li><li><a href="http://terraform.io/">Terraform</a> for IaC</li><li>.Net/Python(we are moving to .Net Core.. excited about running on Graviton. I will cover this topic separately)</li><li>AWS managed services — EventBridge, RDS, and a plethora of other managed services</li></ul><p>↑<a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-6bc494d484cc">Table Of Contents</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c7d865484633" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/why-i-prefer-serverless-c7d865484633">Why I prefer Serverless</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Developer NPS — Why are you not measuring this?]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/the-developer-nps-why-are-you-not-measuring-this-cdf942703aa8?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/cdf942703aa8</guid>
            <category><![CDATA[cto]]></category>
            <category><![CDATA[engineering-culture]]></category>
            <category><![CDATA[engineering-mangement]]></category>
            <category><![CDATA[chief-technology-officer]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Sun, 08 Aug 2021 19:44:42 GMT</pubDate>
            <atom:updated>2021-09-17T00:48:29.361Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Developer NPS — Why are you not measuring this?</h3><p>My first instinct was to write this post completely click-baitey: try this one weird trick to improve the happiness of your engineers :)</p><p>Organizations measure NPS score of their products through standard surveys. Customer support organizations measure tNPS (transactional NPS) and CSAT (customer satisfaction) regularly. Organizationally, Employee Engagement is measured and analyzed.</p><p>There is one additional measurement that most organizations miss out on, that is a must. This is an NPS measurement based on the following question</p><p><strong><em>Would you recommend this code base to a friend?</em></strong></p><p>Why is this NPS score important? It helps leaders understand what the relationship between developers and the code base is. When they start developing, are they scared? Excited? Stressed? My experience is that if you have a code base older than a decade, <strong>you will have an NPS score less than -20. </strong>In other words, horrible. Engineers dread making changes to this code base. In monoliths, it is the proverbial ball of mud. You make a change here, and something else is going to break.</p><p>I use the NPS score — sectioned by teams — to understand where the need for change is the greatest. The NPS score, combined with a code churn score (where is the code base changing the most?) points you the area of the code most in need of reduction of technical debt. (I was going to write about technical debt, but <a href="https://blog.crisp.se/2013/10/11/henrikkniberg/good-and-bad-technical-debt">this blog</a> captures it better than I ever could)</p><p>The most important thing about developer NPS is to track trends and not be too concerned about current scores. I like to measure it every quarter — if the score is not improving gradually, you are either not <a href="https://medium.com/a-technical-leaders-toolbox/establish-a-financial-framework-290ea8540bf6">spending enough on tech modernization </a>or you are focused on the improving the wrong part of your tech stack.</p><p>→<a href="https://quickbooks-engineering.intuit.com/root-cause-analysis-an-essential-systems-skill-be2d0b96d31e">RCA’s — an essential systems skill</a>,↑<a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-table-of-contents-6bc494d484cc">Table Of Contents</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cdf942703aa8" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/the-developer-nps-why-are-you-not-measuring-this-cdf942703aa8">The Developer NPS — Why are you not measuring this?</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Embrance change]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/embrance-change-feab58899a37?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/feab58899a37</guid>
            <category><![CDATA[engineering-mangement]]></category>
            <category><![CDATA[cto]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Sat, 31 Jul 2021 20:29:18 GMT</pubDate>
            <atom:updated>2021-07-31T20:29:18.572Z</atom:updated>
            <content:encoded><![CDATA[<h3>Embrace change</h3><p>Everyone has a plan till they get punched in the mouth — Mike Tyson</p><p>Plans are nothing. Planning is everything — Dwight Eisenhower (?)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*MNo1wpl8YgVHTsWisDQaZA.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@bayleejadegramling?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Baylee Gramling</a> on <a href="https://unsplash.com/s/photos/punch?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>No amount of planning will prevent you getting punched in the mouth. And that is no reason not to plan.</p><p>I wrote previously about <a href="https://medium.com/a-technical-leaders-toolbox/tactics-and-strategy-1ce9d279079b">tactics and strategy</a>. Strategy translates to target states and working backwards from your target state. Tactics (e.g. quarterly plans) ensure that incremental planning is being done to move towards the target state</p><p>However, you will — guaranteed — get punched in the mouth during this process. In fact I will postulate the following ‘Sid’s law’ in the vein of <a href="https://en.wikipedia.org/wiki/Hofstadter%27s_law">Hofstradter’s law</a></p><p><em>“Your plans will get punched in the mouth, despite precautions made because of Sid’s law’</em></p><p>A key employee will leave. A vendor will be 6 months late. Unexpected complications will delay projects. But the answer is not more planning or less planning. The question you should be asking your self is ‘How can I take advantage of this?’</p><p>My experience is that every change, every shock to the system comes with a silver lining.</p><p>A key employee leaving opens up avenues for others to step up. It shines a light on what was hidden thus far. It triggers rethinking about how your teams are organized.</p><p>A delay in implementation triggers new ideas about flaws in the design. One seminal moment I will always remember is when a critical project kept getting delayed (it involved redesigning components that had been written in VB. Microsoft had already sunset the tooling). The delay in the project caused us to rethink what we were doing, and we were able to get rid of the functionality altogether and replatform it.</p><p>Ongoing problems with a Vendor asked us to question ‘would it not be better to bypass the vendor altogether and do direct integrations?’ which is what we ended up doing</p><p>Change is going to happen. You should plan for it. And you should embrace changes even when they force you to replan. They are leading you down a better path. Stop fighting it, and learn from it.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=feab58899a37" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/embrance-change-feab58899a37">Embrance change</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Build an inclusive culture]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/build-an-inclusive-culture-73698792d965?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/73698792d965</guid>
            <category><![CDATA[cto]]></category>
            <category><![CDATA[engineering]]></category>
            <category><![CDATA[diversity-and-inclusion]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Thu, 29 Jul 2021 23:28:01 GMT</pubDate>
            <atom:updated>2021-07-31T17:54:38.054Z</atom:updated>
            <content:encoded><![CDATA[<p>There are a few things in life I consider it ok to be faith based. At the workplace, one is diversity &amp; inclusion. I do not need proof or data that workplace diversity is good. It just is. Having different experiences and backgrounds lends itself to a workplace which is more interesting, more balanced, and lends itself to a greater diversity of ideas. It is more reflective of the diversity of your customers. It will result in better business outcomes without seeking it.</p><p>Who does this well? I will point to Etsy as place which has done a great job with diversity — <a href="https://blog.etsy.com/news/2021/expanding-diversity-equity-and-inclusion-across-etsys-entire-business/">see this article</a>, for example. Intuit, my previous employer, does a <a href="https://www.intuit.com/company/corporate-responsibility/diversity/">fantastic job of inclusion</a>. This is led from the top by the CEO.</p><p>The most common complaint in diversity hiring in engineering is ‘but the pipeline is so poor. I cannot wait a year to find the right candidate’. Well — actually you can, and you should. This was a lesson from Bill Campbell, delivered to me by my manager, Mike, who was lucky enough to be coached by Bill. You have to consider what is more important to you — short or long. If you are playing the long game — and you should be, in most cases — have the courage to put aside some positions for hiring diversity candidates. Refuse to solve for the short by filling it quickly — instead, wait and find the right candidate that will help with your diversity goals.</p><p>This is likely a controversial stance. I believe it is the right stance. And I implement this in my organization, because I am solving for the long.</p><p>Establish a growth mindset, ↑<a href="https://medium.com/a-technical-leaders-toolbox/the-cto-toolbox-table-of-contents-6bc494d484cc">Table Of Contents</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=73698792d965" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/build-an-inclusive-culture-73698792d965">Build an inclusive culture</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Systems optimize for inspection points]]></title>
            <link>https://medium.com/a-technical-leaders-toolbox/systems-optimize-for-inspection-points-c26ef2c15972?source=rss----fe5f447ffd93---4</link>
            <guid isPermaLink="false">https://medium.com/p/c26ef2c15972</guid>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[chief-technology-officer]]></category>
            <dc:creator><![CDATA[Siddharth Ram]]></dc:creator>
            <pubDate>Wed, 14 Jul 2021 23:47:20 GMT</pubDate>
            <atom:updated>2021-07-14T23:47:20.312Z</atom:updated>
            <content:encoded><![CDATA[<p>I recently had a discussion with my peers about engineering productivity and metrics around our <a href="https://medium.com/a-technical-leaders-toolbox/establish-a-financial-framework-290ea8540bf6">fiscal allocations</a>. some concerns about the rate at which certain software defects were being resolved. There was one concern: defects labeled P3/P4 (lower priority) were being closed at a slower and slower rate over time.</p><p>The realization I had was the system was performing exactly as intended. We had inspection points in the system: the inspections were about delivering value to the customers (product roadmap) and ensuring quality attributes (availability/performance/security/product quality) were where we expected it to be. Product quality was assessed by the number of P1 (highest priority) and P2 (2nd highest priority) defects. We were meeting all our SLA’s with P1/P2 defects. Availability was where we needed to be.. just P3/p4 (lower priority defects) were not being fixed at the rate we wanted. The SLA’s were not being met.</p><p>The system was indeed performing as expected: <strong>because a system optimizes for inspection points</strong>. There was no inspection of P3/P4 defects (though we have an SLA for those as well) — decision making on these was local, and inspection was local. On the other hand, all the execs were all over the numbers related to P1/P2, Availability… When a trade off needs to be made, what will a team optimize for? The metrics that are going to be inspected, of course.</p><p>Design your input metrics carefully. Systems will optimize for those. Be ok with letting other things slide. Focus on your critical few metrics, and ensure everyone does a great job on those.</p><p>If you try to optimize for everything, nothing will get optimized.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c26ef2c15972" width="1" height="1" alt=""><hr><p><a href="https://medium.com/a-technical-leaders-toolbox/systems-optimize-for-inspection-points-c26ef2c15972">Systems optimize for inspection points</a> was originally published in <a href="https://medium.com/a-technical-leaders-toolbox">The CTO’s toolbox</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>