<?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 Michal Davidson on Medium]]></title>
        <description><![CDATA[Stories by Michal Davidson on Medium]]></description>
        <link>https://medium.com/@MsArchitect?source=rss-cc4b3fc75378------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*Rvafqe0fNsOeOBcmLomrww.jpeg</url>
            <title>Stories by Michal Davidson on Medium</title>
            <link>https://medium.com/@MsArchitect?source=rss-cc4b3fc75378------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 23:12:52 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@MsArchitect/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[How to Win a Hackathon]]></title>
            <link>https://levelup.gitconnected.com/how-to-win-a-hackathon-ee740c6d47db?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/ee740c6d47db</guid>
            <category><![CDATA[marketing]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[hackathons]]></category>
            <category><![CDATA[software-architecture]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Tue, 04 Feb 2025 16:08:42 GMT</pubDate>
            <atom:updated>2025-02-04T16:08:42.997Z</atom:updated>
            <content:encoded><![CDATA[<p>Do’s and Don’ts for a successful Hackathon POC and Pitch</p><figure><img alt="Desk with many laptops and people working" src="https://cdn-images-1.medium.com/max/1024/1*Ty2cHnNgXWVmDdVub968Cg.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@marvelous?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash">Marvin Meyer</a> on <a href="https://unsplash.com/photos/people-sitting-down-near-table-with-assorted-laptop-computers-SYTO3xs06fU?utm_content=creditCopyText&amp;utm_medium=referral&amp;utm_source=unsplash">Unsplash</a></figcaption></figure><p>I’ll be judging at a company Hackathon next month, and in the run up to that I’ve been mentoring a lot of Hackathon teams. I realized that my advice is relevant to helping you win your Hackathon as well, whether it’s also an internal corporate or academic Hackathon, or a public competition, there are guiding principals that apply everywhere.<br>My list of do’s and don’ts is based on my own experience —in the past few years I’ve won one local and two global Hackathons, been a mentor for Hackathon teams, and judged at a global Hackathon (with the second coming up soon) — so I hope I’m able to provide a well rounded perspective.</p><h4>Tip 1: Complain about everything that bothers you</h4><figure><img alt="woman complaining to man who is looking at laptop screen" src="https://cdn-images-1.medium.com/max/1024/1*D2ZgByyQeDBo_8m0nzg-7w.jpeg" /><figcaption>Photo by <a href="https://www.pexels.com/@keira-burton/">Keira Burton</a> on <a href="https://www.pexels.com/photo/woman-blaming-husband-in-home-quarrel-6624291/">Pexels</a></figcaption></figure><p>Complain- in a group. Yes, really. Meet with some colleagues or friends, and start venting about all the things in your role, company, and product, which are really, really, annoying. Those boring tasks you wish someone else would do instead of you, which waste a ton of time. The stuff that always goes wrong. Those features which don’t work the way they should. The things you wish your product — and products you use — could do. This may feel like a support group but is actually prime ground for brainstorming.</p><p>In Product Management lingo these complaints are “pain points” — and the best innovation comes from solving this pain.</p><p><strong>Do: Identify Real World Problems</strong></p><p><strong>Don’t: Focus on the “How” instead of the “Why”</strong></p><p>I often see teams focusing on a really cool technology that they want to use, or a sophisticated method of using a technique that nobody else has thought of, and when they have the whole product and architecture mapped out, only then try to think <em>why </em>anyone would want to use it.</p><p>Believe me, when your focus is playing around with tech and not solving a real problem — that shows.</p><p>Small caveat here: A Hackathon can be a great opportunity to try out a new technology, and if that’s all you’re in it for — that’s totally legitimate. In fact, for anyone who wants to learn AI, I recommend Hackathons as a great way of playing around with GenAI and gaining hands-on experience with LLMs and the tools and frameworks in that space, such as RAG and AI Agents. But if you also want a shot at <em>winning </em>— the tech alone is not going to be enough.</p><h4>Tip 2: Focus on the “Good Path”</h4><p>The beauty of a Hackathon (and also its weakness) is that the product you develop does <em>not </em>have to work in all scenarios. The gap from an initial Proof-of-Concept (POC) to a product revolves around designing and testing for every single edge case, considering both brown-field and green-field environments, infrastructure changes, day-2 operations, and non functional requirements such as scale. In a Hackathon you <strong>must</strong> ignore all this, or you’ll never be done. In other words, Hackathons are not for perfectionists!</p><p><strong>Do: Your goal in the Hackathon should be to get a demo running from A to Z successfully.</strong></p><p><strong>Don’t: Focus on edge cases and perfection.</strong></p><h4>Tip 3: Define a Doable Scope</h4><p>When it comes to innovation, the sky’s the limit. But when it comes to Hackathons, there <em>is</em> a limit, a very strictly enforced limit, and that’s <em>time.</em> Your development time is typically limited to 24–48 hours.</p><p>Because of this constraint, you must scope your idea very carefully. What is the key selling point of your idea, and what is nice to have, but can wait?</p><p><strong>Do: Define a small and doable proof-of-concept which is achievable in the Hackathon time frame</strong></p><p><strong>Don’t: Add extraneous bells and whistles to your idea.</strong></p><p>Not only will adding extra functionality steal development time and focus, it will also detract from marketing the main idea. You want your idea to be clear and cohesive. Something you can capture in a few sentences, and pitch in a few breaths. (Right, you do get a few minutes to pitch your idea, but you should be aiming to sell the essence of your idea to the judges immediately, in your opening statement.)</p><p>When the extra features are valuable, and more importantly are sellable- but they’re not achievable in the time you have — what you can do is list all these features in your Hackathon pitch, as part of your roadmap. Don’t develop them, don’t demo them, but do list them in your “What’s Next for your Product” slide.</p><h4><strong>Tip 4: Take Short Cuts — A.K.A. Putting the “Hack” in Hackathon</strong></h4><p><strong>Do: Take advantage of every hack possible in order to get your POC up and running within the time frame.</strong></p><p>I mean quick-and-dirty development, ignoring best practices, and even more dramatically- taking advantage of short cuts and workarounds.</p><p>For instance — a team was discussing with me what to do, since they didn’t have API access permissions to a specific data base. For the Hackathon — I suggested they bypass this by copying the data they need to a local data base.</p><p>Another team didn’t have time to develop a pre-processing algorithm for their data, so I suggested they manually label a subset of the data, which was sufficient for the Hackathon demo.</p><p>Particularly when developing an idea using GenAI, you’ll need to take short-cuts. Because LLM results are non-deterministic, and for many use cases also non-predictable, you may achieve the outcome for one use case almost immediately, but it’s likely you won’t get the results you’re looking for when the input changes even slightly. For a Hackathon it’s good enough to select a few use examples that <em>do </em>work, and demo those, leaving out of your demo the scenarios where your LLM is still failing.</p><p>This gap between sometimes-working and usually-working is a major part of the effort involved in extensively researching an idea either pre- or during product development, and it’s one of the reasons organizations are struggling to introduce GenAI, — since for productization the success rate needs to be much, much, higher. Luckily, for a Hackathon POC, partial success is good enough - construct your demo to showcase those successes.</p><p><strong>Don’t: Try to achieve well architected code and end-to-end integration with multiple 3rd party systems.</strong></p><p>I hope I don’t draw flak on this from the architecting community, but a Hackathon is a time for hacking, not for designing beautiful systems.</p><h4><strong>Tip 5: Record your Demo</strong></h4><figure><img alt="Universal Studio movie shutter" src="https://cdn-images-1.medium.com/max/1024/1*stdujLGZklfowu6oc30cnA.jpeg" /><figcaption>Photo by <a href="https://www.pexels.com/@cottonbro/">cottonbro studio</a> on <a href="https://www.pexels.com/photo/black-and-white-labeled-box-3945316/">Pexels</a></figcaption></figure><p><strong>Do: Record a successful Demo for replaying during the Pitch</strong></p><p><strong>Don’t: Rely on a live Demo</strong></p><p>Depending on who you are, my advice to record your demo is either obvious, or blasphemy.</p><p>I’m sorry, I know many of you adore your live demos and feel a pre-recorded demo is cheating. But trust me, live demos are not worth the risk. I’ve sat through so many demos that went wrong in front of the judges- don’t do it!</p><p>And if you’re using GenAI, then remember what I said before about non-deterministic results? When the LLM produces the result you’re happy with- record them for posterity (at least as backup until you’re satisfied the results are reliably reproducible).</p><p>If you’re looking for an authentic and natural flow to your demo, which is difficult to obtain with a pre-recorded and canned message, then my advice is to record the demo <em>without </em>a voiceover or sound track, and during the pitch provide a live running explanation of what the judges are seeing on the screen.</p><h4><strong>Tip 6: Prepare your Pitch in Parallel to your Project</strong></h4><p><strong>Do: At least one team member should dedicate a large chunk (or all) of their time to preparing the Pitch.</strong></p><p><strong>Don’t: Wait till your project and demo is complete to prepare the pitch.</strong></p><p>One thing I can guarantee, and that is you’ll be working until the last minute on developing your project.</p><p>If you leave your pitch to the end- it’s way too late.</p><p>Once the initial functionality and scope of the project are decided upon, send a team member off to design your presentation and practice the pitch. Practice enough times to ensure you can stay within the time limit!</p><p>Judges get tetchy when you don’t respect the rules and your pitch runs over schedule.</p><h4><strong>Tip 6: Dry Run your Pitch</strong></h4><p><strong>Do: Ask for feedback from people with a similar profile to the Hackathon judges</strong></p><p><strong>Don’t: Practice your Pitch only on your team</strong></p><p>I’ve heard many frustrated engineers complain that Hackathons are all about marketing. Sorry folks, but <em>any</em> business is about selling, and the tech world isn’t above that. Besides, learning how to sell your idea is a useful soft skill to have.</p><p>You’ve obviously sold your idea already to your team mates, otherwise you wouldn’t be working on this idea.</p><p>To win the Hackathon- pitch your idea to <em>other </em>people, with as similar a profile to the judges as possible. i.e. if some judges are Tech folks then Tech folks, if there are judges from Product Management then have a dry run with other Product Managers, if business executives then find another business exec who’ll agree to provide feedback. Their feedback is super valuable. Listen and accept any feedback with an open mind, incorporate missing info in your pitch, cut details that are repetitive or unnecessary, and do whatever it takes so your pitch is ultra-clear to someone without your deep tech background.</p><p>I remember when I pitched an idea during a dry run and our mentor told us we’re missing numbers. It only took me an hour to get hold of the numbers for the original product’s infrastructure costs, and show how our Hackathon project would decrease those costs. That definitely helped us win globally- because nothing speaks to Execs like cost savings.</p><h4><strong>Tip 7: Develop your Pitch as if you were pitching a startup idea</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mzD6_4FnkURHHEtqTLh9tg.jpeg" /><figcaption>Photo by <a href="https://www.pexels.com/@rdne/">RDNE Stock project</a> on <a href="http://: https://www.pexels.com/photo/woman-holding-white-paper-with-graphs-7413913/">Pexels</a></figcaption></figure><p><strong>Do: Learn from Startups how to Pitch a successful idea</strong></p><p><strong>Don’t: Spend most of your Pitch explaining the technical complexity of the product- that may impress the audience but won’t sway the judges.</strong></p><p>I couldn’t find much advice online on pitching a hackathon idea. But there is a ton of material available on how to pitch your <em>startup </em>idea. So learn from startup founders.</p><p>Be an entrepreneur on stage, not an engineer. Focus on the Why- the problem you want to solve, and the What- your idea and it’s benefits. Cover in brief also the How- since this touches on the actual tech innovation (which is usually also valued in a Hackathon). I like to also provide a high level timeline or phased development plan- to show your idea is attainable and not a pipe dream.</p><h4><strong>Tip 8: Go for the “Wow!” effect</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*776HYKTqs7DC4AubSCRXKw.jpeg" /><figcaption>Photo by <a href="https://www.pexels.com/@olly/">Andrea Piacquadio</a> on <a href="https://www.pexels.com/photo/cheerful-man-using-laptop-for-video-call-3799821/">Pexels</a></figcaption></figure><p><strong>Do: Ensure your POC includes a usable feature</strong></p><p><strong>Don’t: Leave your project’s impact up to the Judges’ imagination</strong></p><p>As engineers we have a tendency to value technological achievement over the product impact. In a Hackathon this translates into patting ourselves on the back when we conquer the biggest technological hurdle, and then demo-ing <em>that, </em>i.e. proudly parading the incredible technical achievement. To many engineers the UI that can then wrap the tech is ‘trivial’, and ‘obvious’. But by focusing on the effort that went <em>into </em>the project, instead of the value coming <em>out </em>of the project, we’re going to be selling our project short.</p><p>Instead — focus on the user experience which will create the biggest impact.</p><p>Can you now develop a UX monitor displaying the project’s run-time behavior?</p><p>Or did you put in place a platform for automation which can perform something we’re used to seeing done manually?</p><p>Your “Wow!” is dependent on your project, but don’t stop until you’ve created (using any short cuts and hacks necessary) a demo where a regular user (from outside the team) watches and says “Cool!”</p><p>Don’t trust the judges to see the value in your complex tech demo, and figure out for themselves how much potential is in that innovation, which can be potentially translated into a cool feature. Instead — develop and demo the “something cool”. (The fact that there’s complex and innovative tech under the hood is impressive- so touch on that in the “innovation” part of your pitch.)</p><p>Most of the time, getting from your technological achievement to the “Wow!’ is the a matter of minimal effort, say under 10% of the total development effort you invested in the project, and that extra investment is worth its weight in bitcoin.</p><h4>Tip 9: Diversity is the way to go</h4><p><strong>Do: Build a heterogenic team of people with different skill sets and strengths</strong></p><p><strong>Don’t: Build a homogenic team of engineers who all specialize in the same domain</strong></p><p>When I say diversity I don’t mean specifically DEI (although I did participate in an all-women’s Hackathon team once, and it was awesome, and yes — we won!), rather I mean that you should be forming your team with diversity of point-of-view, of background and of skill set. You may all be engineers, but try and find someone who excels at written and visual communication and someone who is confident at pitching an idea under pressure. Also try and combine in your team experts in different tech fields, ideally with at least one UI/UX expert. As a software architect, I admit I may be biased, but I do think someone with architectural and design vision helps as well, especially in order to define the doable scope.</p><h4>Tip 10: Volunteer for the Role you want, not the Role you have</h4><p><strong>Do: Work on skills you want to learn</strong></p><p><strong>Don’t: Stay in your comfort zone</strong></p><p>This is my one piece of advice which isn’t necessarily about winning, rather about helping you gain the most from your participation in a Hackathon.</p><p>While I recommended forming a team with heterogenic skills, I don’t mean formal experience is needed. Perhaps you’re naturally good at marketing/presenting/architecting but you never had the opportunity before? A Hackathon is an amazing opportunity to stretch yourself, learn and experiment with new skills which aren’t part of your day-to-day job description.</p><p>Sometimes these are technical skills such as coding in a new language, or a new technology such as GenAI, and then it’s obvious what you’re gaining from the Hackathon.</p><p>But sometimes you may want to improve your soft skills- such as leadership or marketing. Or you may have been trying to switch roles and missing the experience needed, and this Hackathon could be the perfect opportunity to gain that experience.</p><p>I know many developers interested in transitioning to Product Management, or Software Architecture, or Engineering Management— so volunteer to fill this role in your Hackathon Team. You’ll be amazed at how much you can learn in a few intensive days.</p><p>I often recommend to engineers wishing to become software architects, that they utilize a hackathon to define the architecture of the POC, and the long term architectural vision of the project. This architecture will be part of your portfolio going forward, you can draw it on a whiteboard during an interview when asked about a project, and for internal transfers it’s an especially good way of proving you have the know-how for architecting.</p><p>Sorry ahead for the corniness, but I really believe in this: When you leverage a Hackathon to learn and grow outside your comfort zone, by gaining new experience and new skills — you’ve already won.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PBBLToFpcpDVrViyTa8soA.jpeg" /><figcaption>Photo by <a href="https://www.pexels.com/@ds-stories/">DS stories</a> on <a href="http://: https://www.pexels.com/photo/medals-on-pedestal-on-pink-studio-background-7267588/">Pexels</a></figcaption></figure><p>Let me know if you use these tips and if they help you win your Hackathon!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ee740c6d47db" width="1" height="1" alt=""><hr><p><a href="https://levelup.gitconnected.com/how-to-win-a-hackathon-ee740c6d47db">How to Win a Hackathon</a> was originally published in <a href="https://levelup.gitconnected.com/">Level Up Coding</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why Every Software Architect Needs to Learn GenAI]]></title>
            <link>https://medium.com/data-science/why-every-software-architect-needs-to-learn-genai-c575a669aec0?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/c575a669aec0</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[leadership-development]]></category>
            <category><![CDATA[data-science]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Wed, 08 Jan 2025 12:01:39 GMT</pubDate>
            <atom:updated>2025-01-08T12:01:39.197Z</atom:updated>
            <content:encoded><![CDATA[<h4>You can’t afford to remain an AI-ignoramus, even if your product isn’t using an LLM</h4><p>If you’re a Software Architect, or a Tech Lead, or really anyone senior in tech whose role includes making technical and strategic decisions, and you’re <em>not </em>a Data Scientist or Machine Learning expert, then the likelihood is that Generative AI and Large Language Models (LLMs) were new to you back in 2023.</p><p>AI was certainly new to me.</p><p>We all faced a fork in the road —should we invest the time and effort to learn GenAI or continue on our jolly way?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d5Kbh3C5f7jsjSTqUGfyOA.jpeg" /><figcaption>Image by <a href="https://www.pexels.com/@souvenirpixels/">James Wheeler</a> from <a href="https://medium.com/r?url=https%3A%2F%2Fwww.pexels.com%2Fphoto%2Fphoto-of-pathway-surrounded-by-fir-trees-1578750%2F">Pexels</a>, text added by author</figcaption></figure><p>At first, since the products I’m working on are not currently using AI, I dipped my toes in the water with some high-level introductory training on AI, and then went back to my day job of leading the architecture for programmatic features and products. I reassured myself that we can’t all be experts in everything — the same way I’m not a computer vision expert, I don’t need to be an AI expert — and instead I should remain focused on high level architecture along with my core areas of expertise- cloud and security.</p><p>Did you face a similar decision recently?</p><p>I’m guessing you did.</p><p>I’m here to tell you that if you chose the path of LLM-semi-ignorance, you’re making a huge mistake.</p><p>Luckily for me, I was part of a team that won a global hackathon for our idea around using AI for improving organizational inclusivity, and that kicked off a POC using GenAI. That then led to co-authoring a patent centered around LLM fine-tuning. I was bothered by my ignorance when AI terms flew around me, and started investing more in my own AI ramp-up, including learning from the experts within my organization, and online courses which went beyond the introductory level and into the architecture. Finally it all clicked into place. I’m still not a data scientist, but I can understand and put together a GenAI based architecture. This enabled me to author more patents around GenAI, lead an experimental POC using an LLM and AI agents, and participate in a GenAI hackathon.</p><p>What I learned from this experience is that GenAI is introducing entirely new paradigms, which are diametrically opposed to everything I knew until now. Almost every single fact from my computer science degrees, academic research, and work experience is turned on its head when I’m designing a GenAI system.</p><p>GenAI means solving problems using non-deterministic solutions. We got used to programmatic and deterministic algorithms, allowing us to predict and validate inputs against outputs. That’s gone. Expect different results each time, start thinking about a percentage of success vs absolute correctness.</p><p>GenAI means results are not linear to development effort investment. Some problems are easy to solve with a simple prompt, others require prior data exploration and complex chains of multiple agentic AI workflows, and others require resource heavy fine tuning. This is very different than assessing a requirement, translating it to logical components and being able to provide an initial decomposition and effort assessment. When we use an LLM, in most cases, we <em>have </em>to get our hands dirty and try it out before we can define the software design.</p><p>As a software architect, I’ve begun assessing tradeoffs around using GenAI vs sticking with class programmatic approaches, and then digging deeper to analyze tradeoffs within GenAI — Should we using fine tuning vs. RAG, where is an AI agent needed? Where is further abstraction and multi-agents needed? Where should we combine programmatic pre-processing with an LLM?</p><p>For every single one of these architectural decision points- GenAI understanding is a must.</p><p>I became a Software Architect after over a decade of experience as a Software Engineer, developing code in multiple languages and on multiple tech stacks, from embedded to mobile to SaaS. I understand the nuts and bolts of programmatic code, and even though I’m not writing code anymore myself, I rely on my software development background both for making high level decisions and for delving into the details when necessary. If as tech leaders we don’t ensure that we gain equivalent knowledge and hands-on experience in the field of GenAI, we won’t be able to lead the architecture of modern systems.</p><p>In other words — I realized that I cannot be a good Software Architect, without knowing GenAI. The same way I can’t be a good Software Architect if I don’t understand topics such as algorithms, complexity, scaling; architectures such as client-server, SaaS, relational and non-relational data bases; and other computer science foundations.</p><p>GenAI has become <em>foundational </em>to computer engineering. GenAI is no longer a niche sub-domain that can be abstracted away and left to Subject Matter Experts. GenAI means new paradigms and new ways of thinking about software architecture and design. And I don’t think any Software Architect or Tech Leader can reliably make decisions without having this knowledge.</p><p>It could be that the products and projects you lead will remain AI free. GenAI is not a silver bullet, and we need to ensure we don’t replace straightforward automation with AI when it’s not needed and even detrimental. All the same, we need to be able to at least <em>assess this decision knowledgeably</em>, every time we face it.</p><p>I’m going to end with some positive news for Software Architects — yes we all have to ramp-up and learn AI — but once we do, we’re needed!</p><p>As GenAI based tools become ever more complex, data science and AI expertise is not going to be enough — we need to architect and design these systems taking into account all those other factors we’ve been focused on until now — scale, performance, maintainability, good design and composability — there’s a lot that we can contribute.</p><p>But first we need to ensure we learn the new paradigms as GenAI transforms computer engineering — and make sure we’re equipped to continue to be technical decision makers in this new world.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c575a669aec0" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/why-every-software-architect-needs-to-learn-genai-c575a669aec0">Why Every Software Architect Needs to Learn GenAI</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[My DOs and DON’Ts of Software Architecture]]></title>
            <link>https://itnext.io/my-dos-and-donts-of-software-architecture-63f707571e47?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/63f707571e47</guid>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[software-engineering]]></category>
            <category><![CDATA[technical-analysis]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[agile]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Thu, 26 Dec 2024 14:46:52 GMT</pubDate>
            <atom:updated>2024-12-27T10:55:31.940Z</atom:updated>
            <content:encoded><![CDATA[<h3>Collated from my personal experience as a SW Architect</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*pOP6lmPVrW0uUTc9Wooiqw.jpeg" /><figcaption>Photo by <a href="https://www.pexels.com/@pixabay/">Pixabay</a> on Pexels</figcaption></figure><p>This advice is based on part of a Meetup talk I gave on “Building the Product Right”.</p><p>As the new year approaches, it’s a time of reflection, and this extends into the workplace. So I thought I’d share some of my own personal “do’s” and “don’ts” at work, which I’ve formulated throughout the years. I’ve focused on Software Architecture, since that’s what I’m mainly involved in, but many of these points are relevant to any career track.</p><h4>DO</h4><p><strong>✔ Remind yourself that everyone is mortal — no matter how senior</strong></p><p>Those who know me now will find this hard to believe, but I was born with major fear-of-authority. In fact, as a kid, I used to morph into a blushing whisperer whenever facing the all-mighty school principal. (I’m so happy that trait skipped a generation and my kids will happily roll down the car window and halloo greetings down the block if they see their teacher or principal outside of school. They don’t see why they should be shy, and they’re right!) The way I gradually overcame this in my career is by reminding myself that we’re all just <em>people</em>. You’ll find it easier to overcome stage fright when speaking to folks more senior than yourself — by remembering they all had their own journeys and challenges, and will often be keen to hear your point of view.</p><p>When you’re navigating a meeting or discussion with multiple parties, by remembering we’re all people first, and software architect/engineer/product manager/[insert job role here] second, we can really tune in to each participant’s point of view, to drive a successful outcome that meets everyone’s goals.</p><p><strong>✔ Clarity is Key</strong></p><p>If you don’t understand, chances are others don’t either — so ASK!</p><p>Particularly when I step into a new role, or take on a new product or project, I’ll find myself in a meeting where 1–2 people are talking confidently and at rapid fire speed and the rest are silently nodding along, while to me the terms or three-letter-acronyms are new and unfamiliar.</p><p>It’s <em>very </em>tempting then to keep quiet and nod along too. You’re sure everyone else understands except for you, and you don’t want to betray your ignorance.</p><p>But if you instead push yourself to ask questions and encourage the speakers to explain themselves, you’ll be helping <em>them </em>gain clarity and translate the complex into something communicable, and you’ll be helping <em>everyone else </em>in the room<em> </em>who 9 times out of 10 do not fully understand either.</p><p>Especially when a topic is complex, and someone is an expert, the most critical skill is to be able to translate that complexity into clarity, and I see that as one of the key roles of an architect.</p><p><strong>✔Document Decisions</strong></p><p>If it’s not written — it didn’t happen!</p><p>Back to those meetings, don’t assume that x time later anyone is going to remember the outcome. And let’s say you do remember the outcome and want to follow up, I guarantee there will be someone else who needs reminding, meaning wasted cycles.</p><p>The same goes for POCs, experiments, and technical exploration — the outcome of that effort should be captured in writing.</p><p>So make sure you capture decisions and results in writing. This is not necessarily a document, it could be a Confluence or Wiki page, a presentation deck, or even an email or chat message for less historic decisions. But keep that record somewhere.</p><p>And don’t assume a meeting recording is good enough. Be honest- how many times have you rewatched a meeting recording?</p><p><strong>✔Define an owner</strong></p><p>Collaboration is important, but there’s going to to be a point where you have to make a decision and move on, even if you didn’t succeed in reaching a consensus. That’s when the defined owner should — after taking into account all the tradeoffs raised — make a decision and allow everyone to progress to the next step of making things happen.</p><p>Ownership is also critical so that items don’t slip between the cracks, with each stakeholder assuming the other is working on the item. With a clearly defined owner, they have the responsibility and the motivation to step up to the plate and drive progress.</p><p><strong>✔ Use Architectural Contracts</strong></p><p>Whenever I lead the architecture of a project that crosses<strong> </strong>business units and development groups, meaning there are many owners — I make sure to create some kind of architectural contract.</p><p>When a product or project is developed by a single business unit, product requirements can be good enough for defining the initial requirements, because it’s clear all the technical work will be owned by a single business unit. However when a large scale project includes work spread across business groups, with each providing their own functionality and services/components, then relying on product requirements focused on the end user is not good enough, because a division of responsibility between business units is needed.</p><p>Product Managers are typically not able to define the requirements per technical service/component, since this is a technical split not visible to the end user. This is where I feel a technical ‘contract’ is a must.</p><p>In a technical contract — along with granular written requirements, diagrams are needed too. Component diagrams are obviously the first step, but I’ve often seen aspects of functionality being missed this way, so I like to augment the contract with Sequence Diagrams, since they enable clearly mapping out the division of responsibility between services and components, and thus between groups.</p><p>The lead stakeholders of each business unit (architects and engineering leads) should then review and sign off on the technical contract, to ensure this cross-alignment.</p><p>This may sound obvious, but you’d be surprised how many times multiple groups can move forward, sure they’re aligned with each other, but in practice there are gaps or conflicting visions. A single product needs a single and unified technical vision.</p><h4>DON’T</h4><p>These often correspond and are parallel to the Do’s:</p><p>❌<strong>Don’t blindly trust directives</strong></p><p>Always give yourself room to question and challenge assumptions - even when stated strongly by someone in a leadership position, in a manner that doesn’t invite feedback.</p><p>Don’t ever let anyone take away your mental autonomy. Think, be in tune to your own voice (often based on years of learning and experience), trust in yourself, and learn how to express your own point of view with confidence and clarity. The end decision may be out of your hands (like we said, someone needs to own decisions, and that owner won’t always be you), but do speak up, voice your feedback, and be an intelligent and valuable contributor to discussion.</p><p>❌<strong> Don’t present a problem without a solution(s)</strong></p><p>This is my biggest piece of career advice, and it applies to all career paths.</p><p>It’s very easy to identify a problem, but try not to ever come along and just point out problems (a.k.a criticize, complain). Instead always strategize at least one solution to the problem, and ideally a few alternative solutions, each with their tradeoffs, and present the problems only <em>together with potential solutions</em>. By presenting the path forward, you’re driving change, instead of merely venting. (The solution chosen may be an alternative you didn’t think of, and that’s OK. By presenting at least some solutions you’ll still be kicking off the innovation and thought around how to improve.)</p><p>❌<strong>Don’t assume if everyone is nodding- they understand, and it’s clear</strong></p><p>Like I wrote above — clarity is key. I’ve yet to ask a question that doesn’t trigger other questions from other participants, who until now were nodding away.</p><p>❌<strong>Don’t over-architect</strong></p><p>It’s tempting to develop every architecture to solve for every edge cases, no matter how esoteric. We all want our work to be perfect. But perfectionism hold back progress. And an over-architected design will lead to inflated development time, which often doesn’t pay off in the long term.</p><p>To me this is one of keys of agility, zooming in on what is valuable in each phase, and focusing on that.</p><p>❌<strong>Don’t collaborate without ownership</strong></p><p>If there’s no clear owner, I promise you will see work never happening, or slipping through the cracks. And have you ever worked on a project that gets stuck in a phase and can’t move forward? This is often due to lack of decision making or nobody stepping up to the plate and driving the progress.</p><p>❌<strong>Don’t Go in Circles</strong></p><p>If you feel you’re project is stuck and going in circles, take a step back and identify the root cause. Is it lack of ownership? Is it a wish for consensus that isn’t happening? Are new stakeholders being pulled in necessesitating reopening decisions? Whatever the case, once you understand the root cause you can troubleshoot and drive the process forward.</p><p>❌ <strong>Avoid “WaterGile”- Make sure your Agile is Agile!</strong></p><p>I strongly believe in Agile methodologies, but only when they’re truly Agile.</p><p>Too often I see companies deeply entrenched in Waterfall, attempting to become Agile, while not able to let go of their Waterfall pillars. They keep their set deadlines, set scope, and heavy processes, and layer on top of that some form of Agile methodology- typically Scrum. So you’ll be seeing “Standups” and “Sprints”, “Epics” and “Stories”, but in practice the core mindset remains linear and not iterative, fixed and not flexible. When this happens, you’re adding in the overhead of the Scrum practices, without gaining any of the benefits.</p><p>In other words- it’s the worst of both worlds.</p><p>There are often legitimate reasons to remain with Waterfall — such as when you’re in an industry that requires committed timelines and scope. But if you do make the decision to introduce Agile, make sure you’re committed to truly becoming Agile, and don’t let yourself be weighed down by the combined approach I call “WaterGile”</p><p>Nothing I’ve written is an exact science, and for every rule there’s an exception. All the same, I hope this assortment of advice, based on my own experience in the industry, proves helpful. I’d love to hear your thoughts!</p><p>Wishing you a happy 2025!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=63f707571e47" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/my-dos-and-donts-of-software-architecture-63f707571e47">My DOs and DON’Ts of Software Architecture</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Fixing The Broken Rung]]></title>
            <link>https://code.likeagirl.io/fixing-the-broken-rung-10e4cb539ca6?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/10e4cb539ca6</guid>
            <category><![CDATA[women-in-tech]]></category>
            <category><![CDATA[codelikeagirl]]></category>
            <category><![CDATA[diversity-in-tech]]></category>
            <category><![CDATA[careers]]></category>
            <category><![CDATA[mysogyny]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Wed, 15 Feb 2023 21:23:05 GMT</pubDate>
            <atom:updated>2023-02-15T21:23:05.433Z</atom:updated>
            <content:encoded><![CDATA[<h4>WOMEN IN TECH</h4><p>Overcoming Unconcious Bias In The Tech Workplace</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OPvAJXqcS19NcZ2q9-TWjA.jpeg" /><figcaption>Photo by<a href="https://www.pexels.com/photo/high-angle-photo-of-woman-on-ladder-2467396/"> Samantha Garrote</a> on Pexels</figcaption></figure><p>Have you ever asked for a promotion and been told, “you don’t have the experience” or “you need to prove yourself first”?</p><p>If you’re a woman, I’m going to guess that has happened to you at least once in your career. And if it hasn’t, it may be because you never tried to ask.</p><p>I’ve mentored dozens of young female engineers in the last decade-and-a-half of my career, as well as learned from my career journey and participated in some beautiful forums and meetups for female engineers. A theme that repeats itself constantly is that moment when you look at everything you’ve done, assess where you want to go, and decide, “I’m ready”. Ready for the next step and the next stop on your career path.</p><p>Only to be told- No. You’re Not.</p><p>Because how can we promote you to &lt;X&gt; if you’ve never done &lt;X&gt; before? How can you be a manager if we never saw you manage? How can you <a href="https://medium.com/itnext/so-what-is-a-software-architect-exactly-dbb6406d5747">be an architect</a> if you have never architect a full end-to-end system?</p><p>When I heard from <a href="https://medium.com/u/26471967e154">Moran Weber</a>, Founder of <a href="https://www.womenonstage.net/academy">Women On Stage</a>, about the <a href="https://www.forbes.com/sites/kimelsesser/2019/10/15/new-leanin-study-the-broken-rung-keeping-women-from-management/?sh=253cf577803b">“Broken Rung”</a> study, it all clicked.</p><p>Because once women are allowed to prove themselves, prove that they can lead, whether technically or managerially, they succeed and then they can climb up the rest of that ladder.</p><p>The issue is with that first missing rung. That very first stage of leadership that women often never even get to try.</p><p>Ways to overcome this:</p><ol><li>Gain the experience informally- So you’re not a manager? You can still volunteer to lead meetings and projects. Fill in when your manager is on vacation. Or do you want to be a tech lead? Volunteer to write that design, research that new technology, and present your finding to the group. Write that white paper. You don’t need to wait for the title to try out the work.</li><li>Practice saying, “Give me this opportunity to prove myself. I know I can do it.”</li><li>If all else fails- ask your manager or whoever it was who said “you don’t have the &lt;fill in the gap here&gt; ” for a <strong>detailed </strong>and <strong>specific </strong>list of the experience you are missing, and ideally, ask to work together on a plan for how you can gain this experience.</li></ol><p>And if you are on the other side of that table (or nowadays — that zoom call) —stop before you say “No” to the employee who doesn’t meet the unconscious picture you drew in your head of the perfect candidate. Pause. Think. And maybe, just maybe, give her that chance?</p><p>Because replacing that missing rung will make the world a better place for all of us.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=10e4cb539ca6" width="1" height="1" alt=""><hr><p><a href="https://code.likeagirl.io/fixing-the-broken-rung-10e4cb539ca6">Fixing The Broken Rung</a> was originally published in <a href="https://code.likeagirl.io/">Code Like A Girl</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cloud Security Needs Defense In Depth]]></title>
            <link>https://itnext.io/cloud-security-needs-defense-in-depth-d6de310248d3?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/d6de310248d3</guid>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[cloud]]></category>
            <category><![CDATA[cloud-security]]></category>
            <category><![CDATA[security]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Sun, 12 Feb 2023 23:11:49 GMT</pubDate>
            <atom:updated>2023-02-12T23:11:49.711Z</atom:updated>
            <content:encoded><![CDATA[<h4>CYBERSECURITY / CLOUD SECURITY</h4><h3>A comprehensive overview to Defense-In-Depth in Cloud Security. A.K.A. “Why acronyms are not enough to secure your cloud”</h3><h4>Cloud Security Tooling</h4><figure><img alt="Cloud Security Tools: CSPM, CASB, CWPP, CIEM, CNAPP" src="https://cdn-images-1.medium.com/max/960/1*dTFOyz5NIB5UNP5PuBF_cA.jpeg" /><figcaption>Cloud Security Tooling</figcaption></figure><p>Do you feel overwhelmed by the amount of four-letter acronyms floating around in the Cloud Security space? CWPP, CASB, CSPM, SASE.</p><p>Then comes the point where you need to ask an intelligent question about how these integrate with your old fashioned (i.e. only three letter) on-prem systems such as DLP, IDS and IDP.</p><p>And just when you think you’ve come to terms with all of these must-have cloud security systems, a new acronym comes along- like CNAPP- which has <em>five</em> letters, so it’s obviously newer and shinier.</p><p>It’s a small wonder that with our CISOs running for their dictionaries, the focus of security in the cloud has been distracted from the main challenge- security.</p><p>Monitoring, alerting, automation, analyyics- these tools are all great. Many would even say these systems are essential, with the scale of systems in the cloud, the speed of deployment changes, and the high level of risk. But let’s not get distracted from the basis of designing a secure system, whether on-prem or in the cloud, and that’s defense-in-depth.</p><p>Along with DevSecOps and incorporating security tooling into your agile process, you should also be incorporating security architecture, secure programming and most importantly- a security mindset, during every stage of your cloud journey.</p><p>Some mistake compliance for defense in depth. Because if I fulfill a brain numbingly long checklist of requirements, surely my system is super-duper-secure?</p><p>Well, no. It isn’t. Many compliance checklists do accomplish their job of listing out the typical building blocks required for building a secure system- such as ensuring you are using strong cryptography and not letting outdated (and hacked) protocols stick around in your systems. And that’s great. But if you don’t perform an end-to-end security analysis of your system — there will be gaps. (And I’m intentionally using the word“analysis” and not “review” here, because you will gain so much added value if you move this up to the early design phase, and don’t leave it for a review at the end.)</p><p>You may be using the latest and greatest version of TLS, and MFA, and FIPS-140 compliant cryptographic libraries, and elliptic curve diffie hellman and secure boot. But somewhere between all those different algorithms you’ll be sending keys in the clear, or storing keys protected with a hard coded password stored somewhere else in the clear, or leaving an instance with a public IP accessible for support service.</p><h4>So how can I build a Secure Cloud System?</h4><p>I’m going to <a href="https://medium.com/itnext/cyber-security-is-too-general-38057f2c23bf">refer to my post on Security Domains</a> here. Because by calling out all the domains needed for secure systems, we can ensure that none of these aspects are overlooked in our Cloud Security strategy.</p><figure><img alt="Cloud Security Domains: Compliance, Network &amp; Identity, Reverse Engineering, Monitoring, Cryptography, Security Architecture" src="https://cdn-images-1.medium.com/max/960/1*ERWApA4edYwOMFTQ2zAhUQ.jpeg" /><figcaption>Security Domains</figcaption></figure><p>When it comes to Cloud Security, using this Security Domains breakdown can ensure we cover all the different aspects of security on the cloud. This is also useful for mapping out the various 4-letter cloud security acronyms, and understanding which Security Domain the tools will help with, and where there are still gaps.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*6MRJlMXY9v9CGpIvGpfZdg.jpeg" /><figcaption>Cloud Security Domains</figcaption></figure><h4>1. Cloud Security Compliance</h4><p>The bad: Many standards are struggling with the changes that the cloud brings. Requirements for hardware based security anchors and hardware based unique identities are often either technically infeasible or exhorbitantly expensive in the cloud.</p><p>The good: There are some cloud specific compliance standards, such as ISO 27017.</p><p>Useful advice: When it comes to Privacy, regulations around not processing users’ data outside the region where the user resides is particularly relevant to the cloud. (Think multi-region cloud deployments here)- so pay attention to that.</p><h4>2. Cloud Network Security</h4><p>Please, please, keep in mind <a href="https://medium.com/itnext/what-is-zero-trust-and-why-its-old-news-deed1cb1a2d7">zero trust principles</a> and don’t fully rely on Network Security alone! The cloud network security terms may be new and shiny (VPC anyone?) , but don’t be wooed into thinking that you should be placing all your trust in them. Never forget that one of the key motivations for Zero Trust was the cloud!</p><h4>3. Reverse Engineering</h4><p>All those skills applied to on-prem reverse engineering (A.K.A hacking) can be applied to the cloud as well, just without any possibility of air-gapping your network. Can a cloud system be even more secure than on-prem system? Absolutely. But what happens if somebody, somewhere, misconfigures one instance? You don’t always want to know. According to NSA, <a href="https://media.defense.gov/2020/Jan/22/2002237484/-1/-1/0/CSI-MITIGATING-CLOUD-VULNERABILITIES_20200121.PDF">misconfigurations are still the largest threat to Cloud Security</a>.</p><h4>4. Cloud Monitoring</h4><p>Now this is where most of those cloud security tools come into play. Monitoring has always needed automation, and where more than with the fast-moving and massively scaled cloud?</p><p><a href="https://medium.com/bugbountywriteup/cloud-security-tooling-series-what-the-heck-is-a-cwpp-aa9758f9a339">CWPP</a>, <a href="https://medium.com/bugbountywriteup/cloud-security-tooling-series-what-the-heck-is-a-cspm-8f37f6b1db19">CSPM</a>, and the new and mighty <a href="https://medium.com/@taimurcloud123/cloud-security-tooling-what-does-a-cnapp-do-exactly-7e64db1714a9">CNAPP</a>- it’s time for you to shine. (<a href="https://medium.com/u/6ce09626f2c0">Taimur Ijlal</a> has an excellent series on Cloud Security Tools, which I highly recommend)</p><h4>5. Cryptography in the Cloud</h4><p>Everything you ever learned in cryptography applies to the cloud as well. That’s the beauty of cryptography -it’s based on mathematics and it’s universally true. HOWEVER.</p><p>The bad: Software will never be as good as hardware, at least when it comes to security.</p><p>Trusted Platform Modules (TPMs) are the cornerstone of building a secure system.</p><p>Keys- should be stored within the TPM, and should ideally never leave it.</p><p>Encryption &amp; signing- should be performed within that TPM (using those keys that should never leave)</p><p>And many TPMs do rely on hardware for creating that protection.</p><p>Now a TPM can be build reliant on software. And also the Cloud vendors do provide their own solutions for this. But still, what can I say, it’s limiting.</p><p>Also, security often relies on a unique and tamper proof identity. This identity is typically also reliant on underlying hardware and is created at manufacturing time. Again- working around this in the cloud can be challenging.</p><h4>6. Cloud Security Architecture</h4><p>Cloud Security Architecture is a field which requires (1) cloud knowledge, and (2) security knowledge, and (3) cloud-security knowledge —</p><p>In order to:</p><ol><li>Architect a security system <strong><em>for </em></strong>the cloud</li><li>Architect a cloud system securely (not the same as 1.)</li><li>Perform security reviews and threat modeling on the cloud</li><li>Design &amp; review security protocols — both cloud focused and generic</li></ol><p>I would say the field of Cloud Security Architecture is still in it’s infancy. And we still have a lot of challenges to overcome. Particularly in the multi-cloud, where we usually don’t to rely on security features conveniently provided by cloud providers.</p><p>But that’s what makes it fun!</p><p>PS. What got me thinking about the topics in this post was a <a href="https://youtu.be/wc4vzfZafaQ">Q&amp;A I participated in with NextLabs</a>, who partnered with the NIST to create their Cybersecurity Expert Series on Zero Trust. Check out my interview and share your thoughts and feedback here!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d6de310248d3" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/cloud-security-needs-defense-in-depth-d6de310248d3">Cloud Security Needs Defense In Depth</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[“Cybersecurity” Is Too General]]></title>
            <link>https://itnext.io/cyber-security-is-too-general-38057f2c23bf?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/38057f2c23bf</guid>
            <category><![CDATA[network-security]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[compliance]]></category>
            <category><![CDATA[security]]></category>
            <category><![CDATA[cryptography]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Sun, 01 Jan 2023 23:02:32 GMT</pubDate>
            <atom:updated>2023-01-04T08:13:12.998Z</atom:updated>
            <content:encoded><![CDATA[<h4>CYBERSECURITY</h4><h4>Whether you’re developing your Cybersecurity career, or searching for talent - Security means different things to different people</h4><p>Take Aristotle’s attempt to motivate us with “The more you know, the more you realize you don’t know.” Combine that with imposter syndrome. Toss in the fact that according to popular culture, all high schools contain students who can easily hack into every server under the sun. Shake it all together, and you’re left with the inevitable conclusion that you know <em>nothing </em>about security and have no hope of developing a career as a Cybersecurity expert.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*Gjrk-3kmi744RI02" /><figcaption>Credits: How To Sell Drugs Online (Fast)</figcaption></figure><p>Since I do want to bolster your confidence, I’m going to challenge our concept of what it means to be a cyber security expert.</p><p>The problem, IMHO, is that we look at Security as one big blob.</p><p>“Security.”</p><p>But hold on, nobody nowadays would claim to be a “Computer” expert. There is operating systems, storage, networking, databases — each a separate field.</p><p>And what software engineer writes that she is experienced in “Programming” on her resume? She will list out the languages that she mastered - C, C++, Java, Python, GO, Javascript, etc.</p><p>And most tech conferences, aside from those defined as generalist, will focus on a particular field, framework or language.</p><p>So why when it comes to Cyber Security do we still wrap it all up under one title?</p><p>I’m going to informally define here a map of security fields, each with its own skill set and domain. I hope this will help those of us who love security to feel more confident about what we know vs. what we want to learn vs. what we prefer leaving to someone else to mess around with.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*ULYk8VDdFqa-jMqvPNf00A.jpeg" /><figcaption>Cyber Security Domains</figcaption></figure><h4>1. Compliance</h4><p>I’ll start with compliance because this is what many people think of when they hear security (especially if their product is B2B). You’ll often see compliance grouped together with security as “Security &amp; Compliance” as if they are one happy family.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*EmbSGEnpQAqYQMkzvMyFRw.jpeg" /></figure><p>Please leave with at least this one takeway:</p><p>Compliance is not the same as Security!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/545/1*GtywoH9TXibO85rDXf-dKw.png" /><figcaption>Compliance &lt;&gt; Security</figcaption></figure><p>Compliance is important. Compliance is the best attempt we have to standardize best practices and enforce them. But if we think that compliant systems are secure system — we are making attackers lives much easier.</p><p>Because compliance needs to be generalized by definition, and it’s impossible to apply a cookie cutter template to a bespoke system and not end up with some gaps.</p><p>(I can point out some gaps that commonly appear, such as using secure protocols but not protecting the keys well enough end-to-end, or using MFA but enabling a non-MFA use case for cases when connectivity is down — but that’s a topic for another post)</p><p>Compliance isn’t only a science, it’s also an art. And you’ll often need legal expertise or compliance-domain-specific expertise in order to receive that precious stamp of approval.</p><h4>2. Network Security</h4><p>I’m going to classify as Network Security all aspects of protecting access to a network, and identifying users who request access to resources within that network.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*EzQizAJMqzzrNdaiNE-S8g.jpeg" /></figure><p>Along with pure networking topics such as:</p><ul><li>Gateways (stateless and stateful)</li><li>Firewalls (including Next Generation)</li></ul><p>I’m going to dump in there all those tools and protocols for user identity including:</p><ul><li>Identity and Access Management (IAM)</li><li>RBAC (Role Based Access Control)</li><li>Single Sign On (SSO)</li><li>Multi-Factor Authentication (MFA)</li></ul><p>And also Cloud Networking topics such as:</p><ul><li>VPCs</li><li>Security Groups</li></ul><p>And I’m also going to state that even in<a href="https://medium.com/itnext/what-is-zero-trust-and-why-its-old-news-deed1cb1a2d7"> Zero Trust systems</a> many of these topics are still relevant.</p><p>Now am I terrified about receiving a mass of angry comments about how I shouldn’t be lumping all this together and it’s practically heresy? Yes I am. But remember — it’s still better than lumping it all together under “Cyber Security”. So feel free to make refinements . Maybe one day I’ll be invited to a “Secure Identity” conference and a “Secure Perimeters” conference. That day is not here yet.</p><h4>3. Reverse Engineering</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*GijZOZG4cYG8NBfJk2PKtw.jpeg" /></figure><p>Now we get to the juicy hollywood action.</p><p>Hacking.</p><p>This is what I should be doing all day.</p><p>This is what I should at least <em>know </em>how to do. From age 13.</p><p>I mean how hard can it be to bypass a few goverment security systems to save the day, preferably in less than 24 hours?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*q9FbD5mVTsOX3GxP" /><figcaption>Photo by <a href="https://unsplash.com/ja/@cbpsc1?utm_source=medium&amp;utm_medium=referral">Clint Patterson</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>I actually did learn how to perform reverse engineering. I played around with IDA and Ollydbg with the best of them. For a very short while I found it cool, kinda like being a Ms. Sherlock Holmes, trying to find a password or bypass an identity check to gain access to various applications. But remember how the famous detective pieces together a diary that was shredded, and spends days on it? And the show fast-forwards until he fits the shreds together into something readable? So that’s the part of white-hat-hacking/reverse engineering that is less glamorous and requires more — ahem - patience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5wwCNdgJG6emCHElnPn0zw.jpeg" /><figcaption>Photo by Mike Haw on <a href="https://www.flickr.com/photos/guyfromva/2440475312/">Flickr</a></figcaption></figure><p>I’m here to say that Pen Testing/ Reverse Engineering/ (White hat) Hacking is a super cool field, but</p><p>(a)you need way more discipline and patience than shown in the movies — i.e. no fast-forward button is available</p><p>(b) you CAN work in security even if hacking is NOT your natural calling from birth.</p><p>I think the stereotypes around (b) are what block many women from entering this field, so it’s time to disprove them.</p><h4>4. Monitoring</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*oFP7HMWE35rZglozg9obDw.jpeg" /></figure><p>I’m going to group here:</p><ul><li>Anomaly Detection</li><li>Intruder Detection Systems (IDS)</li><li>Intruder Detection Prevention (IDP)</li></ul><p>These all fall under the category of Analytics/Machine Learning /A.I. — and while fascinating, are really more about applying Data Science &amp; Machine Learning expertise to the security domain, rather than pure security fields.</p><p>In other words — in order to recognize an abnormal user behaviour which may signal some form of malicious use of the system- you need to analyse the user behaviour and compare it to expected, valid, user behaviour, and while the <em>application </em>of this is for security, and the <em>result </em>is improved security, the expertise needed has nothing to do with security and everything to do with DS/ML.</p><h4>5. Cryptography</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*tX8P_rokvVEZOz2YJRLOgQ.jpeg" /></figure><p>Finally!</p><p>To me, cryptography is security at it’s most pure. I love how the security of the world-wide-web rests on deceptively simple arithmetic such as the <a href="https://medium.loopring.io/learning-cryptography-diffie-hellman-key-exchange-discrete-log-problem-cyclic-groups-28703f3d5a48">discrete log problem</a> or the <a href="https://medium.com/coinmonks/integer-factorization-defining-the-limits-of-rsa-cracking-71fc0675bc0e">integer factorization problem</a>.</p><p>My advice to anyone who want to learn more about cyber security is to start by learning cryptography. Not necessarily in depth, but at least the building blocks. There are some great free online introductory courses, and don’t feel you need to be able to implement Open SSL from scratch, but if you understand: asymetric crypto (RSA,diffie hellman, what elliptic curves are at a basic level) , symmetric crypto (AES), signing/verifying, encrypting/decrypting, hashes, and public and private keys- you’re off to a very good start.</p><p>Cryptography then leads into fields such as:</p><ul><li>Embedded security</li><li>Secure Boot</li><li>Key Protection</li><li>Trusted Platform Modules (TPM)</li></ul><p>Because along with the algorithms, we need to generate the keys securely, store them securely, and use them securely.</p><h4>6. Security Architecture</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*vfFr91BLJ9YYd5XD3q5b8w.jpeg" /></figure><p>Security Architecture requires knowledge of (5)Cryptography and usually also aspects of (2)Network Security, but it’s more than that. You take knowledge of the other domains and use it to:</p><ol><li>Architect a security system</li><li>Architect a system securely (not the same as 1.)</li><li>Perform security reviews and threat modeling</li><li>Design &amp; review security protocols such as SSL\TLS and IPSec.</li></ol><p>You <em>do </em>need to understand the hacker mindset, in order to identify where weaknesses may exist in the system and spot vulnerabilites- but that doesn’t mean you need to be able to be a hacker yourself.</p><p>As long as you remember: Compliance &lt;&gt; Security!</p><h4>And that’s it folks</h4><p>I split up Cyber Security into:</p><ol><li>Compliance</li><li>Network Security</li><li>Reverse Engineering</li><li>Monitoring</li><li>Cryptography</li><li>Security Architecture</li></ol><p>You probably would have split up the security domains differently. And that’s OK. My goal here was to:</p><p>(1) Encourage diversity in the field of Cyber Security — it’s a fascinating world full of opportunity- even if you aren’t a hacker.</p><p>(2) Start a dialogue around domains in Cyber Security, to enable us to create communities around specific domains.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=38057f2c23bf" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/cyber-security-is-too-general-38057f2c23bf">“Cybersecurity” Is Too General</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Video Security = The Original ZTNA]]></title>
            <link>https://itnext.io/video-security-the-original-ztna-f51f4e41f9e?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/f51f4e41f9e</guid>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[security]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[zero-trust]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Wed, 21 Dec 2022 08:17:24 GMT</pubDate>
            <atom:updated>2022-12-21T08:17:24.652Z</atom:updated>
            <content:encoded><![CDATA[<h4>SOFTWARE ARCHITECTURE/SECURITY</h4><h4>Video Security implemented a perfect Zero Trust architecture way before Zero Trust was invented.</h4><figure><img alt="Zero Trust Timeline- DVB Conditional Access pre-dates ZTNA" src="https://cdn-images-1.medium.com/max/1024/1*OWYMqWOTzEAe7MCkyYWvwQ.png" /><figcaption>Video Security implementing ZTNA back in 1995</figcaption></figure><p>Most people accept that the Zero Trust approach began in 2003–2004. They’re wrong! Zero Trust Network Architecture (ZTNA) began a decade earlier. Let me explain.</p><p>In <a href="https://medium.com/@MsArchitect/deed1cb1a2d7">Part 1</a> of this series, we covered a beyond-the-buzz-words overview of Zero Trust. Now I will prove according to the very scientific <a href="https://en.wikipedia.org/wiki/Duck_test">Duck test</a>, that Video Security is a Zero Trust solution, and it’s the very first ZTNA around, dating all the way back to the prehistoric times of 1995 when Netflix didn’t exist, viewers were at the mercy of TV broadcasters’ lineups, and the Jericho Forum hadn’t yet uttered the word “perimeter”.</p><p>Why should you care?</p><p>Because one of the huge challenges around implementing a secure ZTNA is lack of industry experience with Zero Trust systems. Or as the <a href="https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-207.pdf">NIST</a> put it:</p><blockquote>“gaps.. arise from a lack of time and experience with mature zero trust deployments”.</blockquote><p>Once we establish that Video Security implements the Zero Trust philosophy, principles, and architecture, we can discuss how to tackle the challenges with a ZTNA, such as BYOD, cloning, and establishing trust in a vast open world.</p><p><strong>Video Security vs. Zero Trust:</strong></p><ol><li><strong>Video security has the same </strong><a href="https://medium.com/@MsArchitect/deed1cb1a2d7"><strong>philosophy</strong></a><strong> as Zero Trust — “Never Trust, Always Verify.”</strong></li></ol><p>The Premier League’s broadcast rights alone are expected to bring in close to $7 billion in 2022 alone. Any viewer who instead watches pirate content is impacting that revenue. 1.2 million UK viewers admitted to watching pirate premier league content in 2021, compared to 2 million legal UK viewers of the premier league. Do the math — Billions of dollars in revenue lost for the Premier league alone.</p><p>Video piracy is often perceived as a victimless crime, and a huge segment of the population engages in viewing of illegaly distributed content without feeling guilty. (Plus pirate broadcasters create video broadcast and video streaming systems with the look and feel of legal systems, so not all piratic viewers are even aware that they are engaging in a crime.)</p><p>So, if you work in video security, you’ve never trusted anyone, least of all the viewers. Zero Trust is nothing new.</p><p><strong>2. Video Security has a Zero Trust </strong><a href="https://medium.com/@MsArchitect/deed1cb1a2d7"><strong>architecture</strong></a></p><p><strong>How is video content protected?</strong></p><p>Video is sent to everyone, either broadcast via sattelite so it reaches every single household with a satellite receiver, broadcast via cable network, or distributed on the public internet. What forces people to pay to view the content?</p><p>The trick is that all video content is <em>encrypted.</em></p><p>There are two main methods for protecting video content. The older one is Conditional Access (CA) used by TV broadcasters. A more modern system is Digital Rights Management (DRM) which is used for protecting streamed content. While both are great use cases for Zero Trust, I will focus here on Broadcast security.</p><figure><img alt="Broadcast video is encrypted" src="https://cdn-images-1.medium.com/max/577/1*oLR7sm4Pe8SrME_JdIosHg.png" /></figure><p>The broadcast content is distributed and the encrypted content, along with the keys used to encrypt it are shared — Definitely no trust in the network here!</p><figure><img alt="(Paying) Viewers are registered with the Control Plane" src="https://cdn-images-1.medium.com/max/565/1*ksMfchrtSgZbcc5vVWmo4Q.png" /></figure><p>The video control plane registers all user subscriptions. Every user will typically have a different set of subscriptions (e.g.: sport, movie channels, news, kids, VOD) — Remember Zero Trust’s “Least Privilege” principle? We don’t let <em>all </em>users access <em>all</em> content.</p><figure><img alt="Devices need to be registered with the Control Plane too" src="https://cdn-images-1.medium.com/max/526/1*mprT4iyfs6SDkodm4Kg5wA.png" /></figure><p>It’s not enough that the user registered in the control plane. Every <strong>device </strong>also needs to register — Like in Zero Trust, the Device needs to be authenticated along with the User.</p><p>This device identity must be (1) unique (2) secure (3) clone proof. <br> (We’ll discuss preventing cloning in more detail in part 3 of this series.)</p><figure><img alt="Keys are sent encrypted" src="https://cdn-images-1.medium.com/max/565/1*ksMfchrtSgZbcc5vVWmo4Q.png" /></figure><p>The Control Plane distributes entitlement keys used to decrypt the (encrypted) content-keys that were sent with the content.</p><p>The entitlement keys are also uniquely encrypted, per device. (Paranoid much, anyone?)</p><p>Only a valid device, for a valid user, with a valid subscription- can access the content key and decrypt the content.</p><p>When we compare the 2 architectures side by side, we can see the similarities. (In part 3 we’ll get into this in more depth, to see how we can improve ZTNA system architectures)</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*xF03sUgF3UvaLpq9d3YICQ.jpeg" /></figure><p><strong>3. Video Security has the same </strong><a href="https://medium.com/@MsArchitect/deed1cb1a2d7"><strong>principles </strong></a><strong>as Zero Trust</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*8_lnrK6qYhIo6Edgv-fzag.jpeg" /><figcaption>Zero Trust Principles (Photos by: <a href="https://unsplash.com/@brett_jordan?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Brett Jordan</a>, <a href="https://unsplash.com/@mengmengniu?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">niu niu</a>, <a href="https://unsplash.com/@raysontjr?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Rayson Tan</a> &amp; <a href="https://unsplash.com/@jdent?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Jason Dent</a> on <a href="http://unsplash.com/">Unsplash</a>)</figcaption></figure><p><strong>Zero Trust Principles in Video Security:</strong></p><ol><li>User + Device Identity: Both viewers and their devices are identified by the Control Plane</li><li>Least Privilege- Users can only watch content they purchased. They can’t share it, copy it, or record it unless they have special permissions.</li><li>Least Exposure — the keys that are used to decrypt the content, are changed every key period. Those keys, even if exposed are not very useful. The entitlement keys used to protect those keys are stored in a Trusted Computing Base (TCB). The user cannot access the keys used to decrypt the content.</li><li>Least Access- this is implemented in the broadcaster’s control plane, with string segmentation between content, user permissions, user keys and content keys (however this is not a defined part of the DVB protocol).</li><li>Cohesion- viewing activity and device activity is constantly monitored for piratic behaviour.</li></ol><p>E.g. If you are watching TV 24/7- are you:</p><p>(a) really really bored</p><p>(b) leaving your TV on for background noise</p><p>(c) a pirate sharing the content?</p><p>I can’t share more than that, but there are a lot of analytics going on here and, even back in the 90s, prior to the widespread usage of A.I/Machine Learning, video broadcasters were analyzing viewer behaviour to identify pirate distributors.</p><p>To summarize, <strong>all the principles of Zero Trust are implemented in Video Security systems too</strong>.</p><figure><img alt="Video Security Principles = Zero Trust Principles" src="https://cdn-images-1.medium.com/max/960/1*pY_8ygSJhLRAp0pzcS-qxA.jpeg" /></figure><p>Back to the <a href="https://en.wikipedia.org/wiki/Duck_test">Duck Test</a>, if a Video Security system looks like a ZTNA, runs like a ZTNA and sounds like a ZTNA, then it probably is a ZTNA.</p><p><em>Coming up in Part 3: How Video Security Teaches Us To Build A Better ZTNA- on Cloning, BYOD and trust in an unstrustworthy world.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f51f4e41f9e" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/video-security-the-original-ztna-f51f4e41f9e">Video Security = The Original ZTNA</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[What Is Zero Trust, And Why It’s Old News]]></title>
            <link>https://itnext.io/what-is-zero-trust-and-why-its-old-news-deed1cb1a2d7?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/deed1cb1a2d7</guid>
            <category><![CDATA[software-architecture]]></category>
            <category><![CDATA[security]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[cybersecurity]]></category>
            <category><![CDATA[zero-trust]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Wed, 21 Dec 2022 08:16:16 GMT</pubDate>
            <atom:updated>2022-12-22T20:36:18.022Z</atom:updated>
            <content:encoded><![CDATA[<h4>SOFTWARE ARCHITECTURE/ SECURITY</h4><h4>A speedy technical overview of ZTNA, beyond the buzz words</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FI623dZ0DljA_PloNemtUw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@phillshaw?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Phil Shaw</a> on <a href="https://unsplash.com/s/photos/clone-trooper?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p>Are you looking for a technical overview of Zero Trust Network Architecture, with actual concrete explanations instead of endless buzz words and marketing slogans?</p><p>We will cover that here, and then dive deeper, when I prove that the Zero Trust paradigms, principles, and even architecture pre-date their accepted history. We will travel back in time to 1995 and see how the Video Security industry created a zero trust model for Pay TV which can still be learned from today, to design more secure Zero Trust systems and tackle the common security challenges facing us currently.</p><p>The standard history you will see for Zero Trust shows the first paper with this shift towards perimeter-less security appearing in Jericho Forum’s work on “De-Perimeterization” in the years 2003–2004.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hd8sLpWdUcfS_nQ8dehjUw.png" /><figcaption>Zero Trust Timeline</figcaption></figure><p>In <a href="https://medium.com/p/f51f4e41f9e/">part 2</a> of this series, I will prove that the Zero Trust approach goes back at least to 1995, with the advent of Pay-TV.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*OWYMqWOTzEAe7MCkyYWvwQ.png" /><figcaption>Conditional Access in 1995? What’s that got to do with Zero Trust?</figcaption></figure><p>But before we get finicky about dates, what on earth is this Zero Trust that I keep going on about?</p><p>To understand Zero Trust, we need to understand what existed <em>before</em> Zero Trust came along, and what Zero Trust <em>changed</em>.</p><p><strong>The Pre-Zero-Trust Security Model</strong></p><p>Once upon a time, in the Enterprise industry, it was a wonderful world. We had trust. We had faith.</p><p>Our users were company employees, they got a salary from the company, they wanted to continue to get paid by the company. Money makes the world go round yada-yada. So obviously, we developers and system designer and architects and security researchers could trust them.</p><p>The challenge was the big bad wolf outside. The competition. Hackers. Enemy governments. We called them all “attackers”, and in Threat Modeling we called them “Malicious Mallory” and “Eavesdropping Eve”.</p><p>We erected fences around our systems, in the form of stateless and stateful firewalls and gateways and DMZs.</p><p>Some organizations took it a step further and cut off contact with the outside world all together, using air-gapped networks with no connection to the public internet.</p><p>Our whole focus was on making sure only the “good guys” could enter our system. We focused on Access Control and Identity and Access Management (IAM); identifying and authenticating users, and then authorizing them with access to the system. We adapted protocols such as LDAP, and later SAML and OAuth 2.0/OIDC.</p><p>This model, with the focus on network security, is often compared to a castle, surrounded by a moat, with guards letting down a drawbridge only for authorized visitors to enter and leave. But I see it more as traffic spikes at the entrance to parking lots, because entry is controlled — but within the parking lots is a free-for-all and <em>everyone</em> can exit freely, taking <em>anything</em> with them as they go.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YgluDRP1PwpLRsiv6t6-qA.jpeg" /><figcaption>Photo by <a href="https://www.instagram.com/baruch.photographer/">Baruch Wisemon</a></figcaption></figure><p><strong>Zero Trust Motivations</strong></p><p>This brings us to the first drawback of the network security model, if (or more accurately –‘when’) a malicious attacker does gain access, they can often escalate privileges and use lateral movement to access the most privileged areas of the system, and once they’ve done that, they can also easily send that information out of the system. Add on that the fact that in 2013 every organization’s bubble burst, when Snowden released classified information and shattered our illusions about how much trust we should have in <em>all</em> our employees.</p><p>Along with that came the Cloud — those whisps in the sky (or VMs on Cloud Service Providers (CSPs))- which meant enterprises needed to worry about their egress and ingress traffic, securing their Virtual Private Clouds (VPC), and of-course, enterprises don’t want to put all their eggs in one basket, so along comes the Multi-Cloud and the fun tasks of securing CSP to CSP traffic.</p><p>And don’t forget Covid- it’s 2019 and we all start working from home.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NyZAPtPdLHnTSKhZIz9FuQ.jpeg" /><figcaption>Working from home- in my dreams.. (Photo by <a href="https://unsplash.com/@instagramfotografin?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Peggy Anke</a> on <a href="https://unsplash.com/s/photos/digital-nomad?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a>)</figcaption></figure><p>OK, JK. This was my Covid Working-from-home experience:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v1STSok4Lbz9HhPYieYHwg.jpeg" /><figcaption>Working from home- in reality.. (Photo by <a href="https://unsplash.com/es/@charlesdeluvio?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">charlesdeluvio</a> on <a href="https://unsplash.com/s/photos/working-from-home?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a>)</figcaption></figure><p>But either way, you’re not in the office. You may be connecting to your organization’s servers via VPN, but have you ever considered the security of your home router and Wifi? And are you using the PC your organization provided or a BYOD mobile device?</p><p>So let’s summarize:</p><p><strong>3 of the main motivations for Zero Trust :</strong></p><p>1. Vulnerabilities in the Network Security model</p><p>2. Public Cloud usage</p><p>3. Hybrid workforces and BYOD</p><p>Now we are finally ready to tackle the momentous question of — what is Zero Trust? Thank you for your patience!</p><p><strong>Zero Trust Definition</strong></p><p>The short answer is the <strong>Zero Trust is “Perimeter-Less” security</strong>.</p><p>A slightly longer formalized answer, kindly provided by the NIST in their <a href="https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-207.pdf">Special Publication 800–207</a> : <em>“Zero trust (ZT) is the term for an evolving set of cybersecurity paradigms that move defenses from static, network-based perimeters to focus on users, assets, and resources.”</em></p><p>I prefer <a href="https://www.paloaltonetworks.com/zero-trust#:~:text=What%20is%20Zero%20Trust%3F,stage%20of%20a%20digital%20interaction.">Palo Alto Networks’s definition</a>:</p><p><em>“Zero Trust is a strategic approach to cybersecurity that secures an organization by eliminating implicit trust and continuously validating every stage of a digital interaction.”</em></p><p>The philosophy here is to no longer have any <em>implicit</em> trust, A.K.A no default, unjustified, trust. In other words — be paranoid. Guilty until proven innocent.</p><p><strong>Zero Trust Principles</strong></p><p>Trust stops being a Boolean yes/no. Instead ZTNA introduces “trust levels” — to what extent do I trust user X on device Y doing action Z at time T. Notice the extra limitations here- this is the ZTNA principal “Limitation of privileges” -which always existed in the security domain, but now gained more traction. Along with the Network Security IAM approach to authenticate and authorize users, ZTNA requires authenticating and authorizing the user’s <em>device</em>. Because perhaps the user’s credentials were stolen? Or perhaps the user is valid but their device is not secure? We also limit the access time. Permanent access keys are a big no-no in the ZTNA model. And authorization models should use time based expiration and renewal requirements and not rely on revocation for identifying malicious actors (for good reason, since it’s incredibly challenging to successfully implement revocation designs such as CRL and OCSP described by many security protocols).</p><p>Let’s summarize the Zero Trust principles:</p><p>1. Every user, device and network flow is authenticated and authorized — continuously.</p><p>2. Least privilege- Restricting the access and permissions users and entities have to only the resources they need.</p><p>3. Least Access- Don’t give any entity access to any service they don’t need to access. This is also known as micro-segmentation, since we divide the network into segments.</p><p><em>It’s not the all-or-nothing of network access. There is no- you are in the company intranet, you used our VPN, you signed in- have fun. Compare the spread of germs in a wide open populated stadium, to the spread of germs when people are divided into “capsules”. (Post-covid we are all epidemiology experts)</em></p><p>4. Least Exposure- Here the focus is on <em>what</em> we our trying to protect. — our most precious data, our most sensitive processes.</p><p><em>Typically this will be data and not logic/code/intellectual property — Expose this data as little as possible and ideally use strong cryptography to protect.</em></p><p>5. Cohesion -the entire system is being monitored and analyzed continuously.</p><p><em>All entities, all system behaviors –analyze it all, and when an anomaly is identified the system should notices. This will always need automation and analytics, and usually also machine learning, in order to identify valid vs. suspicious user behavior. Cohesion is often the most challenging aspect to introduce into legacy system.</em></p><p><strong>ZTNA Reference Architecture</strong></p><p>The NIST <a href="https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-207.pdf">proposed a reference architecture in 2020</a>. Their goal was to create a set of logical components, and a common language for ZTNA system architectures.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/835/1*WDjaKd4qC3zfIfP8bljzNg.png" /></figure><p>We see here the authentication of both the user and their device, with the Policy Enforcement Point (PEP) being per Resource, and not a single point of entry into the full system. The control plane based Policy Engine will establish the trust level based on external inputs (such as analytics) along with internal business logic, and if access is to be authorized, the Policy Engine will provide access, often using a cryptographic and time limited access token, which the PEP can use to access the resource.</p><p><strong>What’s Missing? A.K.A the Parallel Universe</strong></p><p>So we have a Zero Trust philosophy, Zero Trust principles, and even a Zero Trust Reference Architecture.</p><p>Nonetheless, NIST highlight a major challenge in designing and building ZTNA systems. The model is new, there are few systems that have fully adapted this model, so we’re learning as we go along.</p><p>And that’s where I disagree with them.</p><p>Because there is another industry where we <em>never</em> trusted our users.</p><p>There is an industry where we were <em>always</em> paranoid.</p><p>Where we always assumed that if viewers could easily access content without paying, they would. And we always worried about a dedicated subset who was eager to to share content in reward for $$$.</p><p>And that’s the Video Security industry.</p><p>Coming up in <a href="https://medium.com/p/f51f4e41f9e/">Part 2 - <strong>Why Video Security is the original Zero Trust system</strong></a><strong>.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=deed1cb1a2d7" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/what-is-zero-trust-and-why-its-old-news-deed1cb1a2d7">What Is Zero Trust, And Why It’s Old News</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[So what is a Software Architect exactly?]]></title>
            <link>https://itnext.io/so-what-is-a-software-architect-exactly-dbb6406d5747?source=rss-cc4b3fc75378------2</link>
            <guid isPermaLink="false">https://medium.com/p/dbb6406d5747</guid>
            <category><![CDATA[career-paths]]></category>
            <category><![CDATA[design-systems]]></category>
            <category><![CDATA[system-architecture]]></category>
            <category><![CDATA[software-architecture]]></category>
            <dc:creator><![CDATA[Michal Davidson]]></dc:creator>
            <pubDate>Sun, 12 Dec 2021 20:13:15 GMT</pubDate>
            <atom:updated>2022-12-22T07:28:24.932Z</atom:updated>
            <content:encoded><![CDATA[<h4><strong>SOFTWARE ARCHITECTURE</strong></h4><h3><strong>So You Want To Be A Software Architect?</strong></h3><p>What it means to be Software Architect and what skills are needed for you to succeed</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*z4ez9RrBLf-OyjucUtlHPw.jpeg" /><figcaption>Photo by <a href="https://unsplash.com/@thisisengineering?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">ThisisEngineering RAEng</a> on <a href="https://unsplash.com/s/photos/female-engineer?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></figcaption></figure><p><strong>First of all- What i<em>s</em> a Software Architect?</strong></p><p>I get asked this a lot. (Mainly by non-programmers, e.g. my doctor, when he wants to know why I can’t just say “programmer” when he asks me my profession, and isn’t an architect someone who designs buildings?).</p><p>I don’t think an architect is just another step on the technical “individual contributor” track. I don’t think it’s “you’ve worked X years as a developer, you don’t want to/aren’t good at managing, but are a very good developer, so we will pay you more and give you a fancy title so that you feel valued”. I also don’t think it’s the same as being a Tech Lead or Principal Engineer.</p><p>I decided early on in my career, after working a couple of years as a developer, that I wanted to be an architect one day. I worked closely with an architect, saw what he did, and thought “Hey, I’d be good at that” and “Hey, I want to do that.”</p><p>It took me a while to get there. Because the career path for becoming an architect is not straightforward. Now I’m a person who likes to google EVERYTHING. But I couldn’t find much advice online on how to become an architect, and what an architect actually is. In fact, I found a lot of titles that mean different things in different places, and many peers and managers telling me that I should be happy with what I have and stay in software engineering.</p><p>It took a few more job changes, a lot more learning, and also lots of luck and amazing people who believed in me and gave me this opportunity, until I reached my dream job.</p><p>So here goes with my personal take on this fuzzy, chameleon like career.</p><p>The most important trait for an architect, A.K.A <strong>Superpower #1.</strong></p><h4>Superpower #1: <strong>See the big picture</strong></h4><p>This is the most important trait. Technology is growing more and more complex. Systems are now composed of tens to hundreds of components, services and microservices. Someone needs to have the technical ability to comprehend that huge system, and then drill down into the details as necessary.</p><p>Does that mean an architect needs to know <strong>everything</strong> about the system?</p><p>That’s a great ambition to have, but pragmatically what can work just as well, depending on the scale of the system, is to supplement that big picture knowledge with the knowledge <strong>where</strong> to look or <strong>whom</strong> to ask, for details which you don’t know yet, along with the ability to comprehend that knowledge and incorporate it into your ever-growing knowledge base.</p><p>(I do like to bring to the table very in depth knowledge of at least some parts of the system, so that it’s not a constant refrain of “I’ll look it up”; nonetheless every role has a learning curve, as you build up more in depth knowledge of the entire system)</p><p>This brings me to <strong>Superpower #2</strong>, which is really more of a passion and a pre-requisite than a skill-</p><h4><strong>Superpower #2: Love Learning</strong></h4><p>You have to be hungry to learn more, every single day. New systems, new technologies, new methodologies. And you’ve got to be an auto-didact, nobody is going to spoon feed you this info. You’re going to be googling and digging around obscure old document databases and reading ancient standards and specs, and calling/chatting/emailing in what feels like a wild goose chase to capture an esoteric fact from a developer who left the company ten years ago. If you don’t love learning, and do it very well on your own, you will make a terrible architect.</p><p>Why does an architect need to know all this, what added value is a random technical person knowing the whole system bringing to the company?</p><p><strong>Architecture-</strong> fitting all the puzzle pieces together so that it all works. Taking a new feature or a brand new system and breaking it down into components/services/microservices. Writing a high level architecture, and/or design, and/or spec, and/or requirements (the process and the document names vary widely between organizations)</p><p>Doesn’t this contradict Agile? It could. But first of all- many companies still use a mishmash of agile and waterfall, and second of all- a big challenge to Agile is scale. In my experience an architect can bring a lot of value to the table with an initial HLA, breaking the product/project/feature/system down into user stories before multiple development teams take the user stories and run with them. And, again, in my experience- many agile teams postpone documentations- forever. Despite it being part of the Definition of Done.</p><p><strong>Non-functional requirements-</strong> scalability, performance, resilience, security — these often get pushed to low priority when the customer is measuring the system according to these criteria- it’s an architect’s job to think about them.</p><p><strong>Design- </strong>This really depends on the role. Sometimes the team and the tech lead will be making these decisions, and that’s great. Sometimes there is a blue-print we want all teams to follow, for reusability and consistency, and then the architect needs to be more involved in defining the blue print. Sometimes an architect will be more dedicated to specific components, and will be the one making the design decisions.</p><p>What I’m seeing is many times an architect with the title “Software Architect” is more involved in design, and a “System architect” less so, but this really differs per company.</p><p><strong>Communication</strong>- constant communication with the engineering teams — developers, tech leaders, QA, other tech teams and architects for integration.</p><p>Now, communication sounds basic, but often very skilled technical people find it challenging to translate their knowledge into clear understandable info.</p><p>So that’s <strong>Superpower #3.</strong></p><h4><strong>Superpower #3: Communicate with Clarity</strong></h4><p>Communication may be external- an architect will often explain to customers what the product does? And more important- what the product <strong>doesn’t do</strong> but <strong><em>could</em> do</strong>- an architect often defines new functionality and features based on either a product roadmap or customer requests. Translating that vision into the nuts and bolts will often be the role of either a <strong>Solution Architect</strong>, <strong>System Architect</strong>, or <strong>Product Architect</strong>. (again, depending on the company- sensing a pattern here?)</p><p>Communication may be within the company, but still external to your organization — with executive management, product and sales organization- defining together the roadmap, the risks of certain features, the need for other features, rough estimations (absolute estimations with commitments should come from the engineering teams, but many times a estimation is needed on the spot). For rapid decision making you often need one person in the room instead of a collection of managers and tech leads.</p><p>The ability to write documents/wikis/confluence pages/whitepapers/whatever documentation style your company uses- is an important soft skill to have, along with the ability to communicate visually with presentations, UML and whiteboards.</p><p>The rest is a bonus- and depends on your own skills and passions, and what your role needs- innovation, research skills, technology ninja super skills, domain specific knowledge (for instance, I love security and cryptography, while a different architect could have a strong knowledge in machine learning or computer vision). Each architect will usually bring a unique set of added value to the table.</p><p>How is this different from being a Tech Lead?</p><p>I got asked this today, and this question is what triggered this post.</p><p>I think one key difference is the scale- architects need to see the big picture of the entire system. A tech lead is usually focused on one dev team of 5–10 developers (2 pizza pie theory anyone?), which means focused on one to a handful of components/services/microservices.</p><p>Another key difference is interactions- the tech lead does need to lead the hands-on development and be a hands-on developer herself. She’s going to be a lot more hands on with the code than an architect typically will be. There are hands-on software architects (usually these are architects who are more focused on one to a handful of components. Because once you go very large scale you will have less of the luxury of time to master dozens of codebases.) Here I agree that the difference between an architect and a tech lead gets fuzzier. I do think the tech lead will still be more focused on task management and possibly people management, while an architect will be more focused on the product vision and documentation. Design, technology choices and processes for development (i.e. CI/CD, TDD etc) can often be done either by the tech lead or the software architect. This often depends on the defined role, the company culture, and the strengths and passions of the architect and tech lead.</p><p>Now if you were super focused and actually read inside the parentheses you may be shuddering at my suggestion that not all architects are hands on. How can any architect be any good technically if they are not writing code themselves?</p><p>Yes <strong>Superpower #4</strong> is important.</p><h4>Superpower #4: Tech Expertise</h4><p>You have to be <strong>experienced in technology and programming</strong>, and algorithms, and logic, and system design, and whatever else your system is using. But I see that as a baseline for any career in Tech Engineering, and not an architect specific super skill. I strongly believe that if architects responsible for large scale systems spend the majority of their time writing code, they won’t have enough time left for all the other things an architect should be doing. (and I’m not talking about the 1% of architects who are above nature and sometimes program a new operating sleep in their sleep between their day job, chess club and breakfast)</p><blockquote>“So why don’t you code?”</blockquote><blockquote>“Do you miss coding?”</blockquote><blockquote>“Do you want to go back to coding?”</blockquote><p>At risk of being expelled from the software-developer-sisterhood, my answer is going to have to be no.</p><p>I hope that I am good at developing code. I studied programming in high school and Computer Science in college and enjoyed it a lot. I had a fun software engineering career. But I’m good at a lot of other things too, besides developing code. What I love about being an architect is that I build on a lot of different super skills. I bring to the table a combination of abilities, and in my opinion that combination is quite rare. I enjoy using this wide variety of skills in my day-to-day work. I used to love programming too, but I love what I do now even more.</p><p>Because whatever our title, my strong belief is that in order to find success and happiness in our careers, we should follow what we enjoy doing (our passions), what we are good at (our strengths) and then if we find market demand for that (as most of us do in Tech) we are extremely lucky!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/336/1*QfAOXEoncpu3S9xb8Tq2wg.png" /></figure><p>And that’s why I’m a software/system/product architect. I love to create something out of nothing (especially if it’s a cool, useful, innovative something), I love to interact with dozens of people and teams every day, be constantly learning, leading, and creating a broad impact, I’m good at understanding complex tech, researching, and communicating, and I’m lucky enough to have a career where I can do this.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dbb6406d5747" width="1" height="1" alt=""><hr><p><a href="https://itnext.io/so-what-is-a-software-architect-exactly-dbb6406d5747">So what is a Software Architect exactly?</a> was originally published in <a href="https://itnext.io/">ITNEXT</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>