<?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 Randall Abele on Medium]]></title>
        <description><![CDATA[Stories by Randall Abele on Medium]]></description>
        <link>https://medium.com/@randall.abele?source=rss-d96b66cf1627------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*aqyZEe6438P_FQvZleAuPA.png</url>
            <title>Stories by Randall Abele on Medium</title>
            <link>https://medium.com/@randall.abele?source=rss-d96b66cf1627------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 20:44:30 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@randall.abele/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[Coding agents have no idea what customers want — Revelica]]></title>
            <link>https://medium.com/@randall.abele/coding-agents-have-no-idea-what-customers-want-revelica-cdc1ff594602?source=rss-d96b66cf1627------2</link>
            <guid isPermaLink="false">https://medium.com/p/cdc1ff594602</guid>
            <category><![CDATA[product-discovery]]></category>
            <category><![CDATA[ai-agent]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[user-story-mapping]]></category>
            <category><![CDATA[product-management]]></category>
            <dc:creator><![CDATA[Randall Abele]]></dc:creator>
            <pubDate>Wed, 11 Mar 2026 00:00:24 GMT</pubDate>
            <atom:updated>2026-05-20T20:04:56.300Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*dM0pDGJrBhZoistw.jpg" /></figure><p>Andrew Ng recently pointed out <a href="https://www.deeplearning.ai/the-batch/ai-product-managers-will-be-in-demand">PMs are becoming a bottleneck</a> because engineers are enjoying a massive productivity boost from AI. YCombinator also noticed this problem which is why their 2026 Request for Startups <a href="https://www.ycombinator.com/rfs#cursor-for-product-managers">asked for a Cursor for Product Managers</a>.</p><p>If we speed up building without speeding up discovery, then we’ll be collectively shipping more stuff people don’t want.</p><h3>Discovery is inevitable, but it only helps you avoid risk if you do it before shipping</h3><p>The planning process is still alive and well, but now it’s better accomplished with an agent than sitting there staring at a ticket screen. Planning with an agent is the only path fast enough to match the pace of AI delivery. You’d be crazy not to use this newfound power. I spend more time iterating and doing ad-hoc discovery research with an agent than I do actually implementing now. By a lot.</p><p>The problem is that the discovery work you do on the fly has none of the structure that helps us avert bias. It’s not cited or recorded in a way that can withstand scrutiny, and is fraught with dangerous assumptions that we (both humans and agents) make when we approach it that way.</p><p>Every time we ship a new innovation, we inevitably discover whether it will be valuable, profitable, and usable. Product discovery is the art of doing that quickly in advance. You might argue that AI delivery speed makes “test in production” a more practical option, and that’s true to a point. But it ignores an important bias: the first idea is almost never the best one in terms of customer value, enhanced margin, or competitive differentiation. You can’t ask your customers to use 17 variants of a feature you just shipped and tell you which one they’d buy! So what actually happens is you ship an imperfect version of what could have been, and it becomes much more difficult to fix that once you have v1 in the hands of your users. What good is a hard-earned customer opportunity if your solution is dogshit? You might just give up on the whole idea when you were actually on to something valuable. This is why we have product discovery in the first place. It lets us iterate and compare ideas in an evidence driven way so we’re not leaving money on the table.</p><p>The answer isn’t to slow down the delivery. It’s to make discovery fast enough so you can learn from experiments instead of your customers and the market at large.</p><h3>Your coding agent doesn’t need to be persuaded</h3><p>PRDs exist because human engineers need narrative context to understand intent. You write a persuasive document that argues for why something should be built and gives enough context to make good decisions along the way.</p><p>Your coding agent doesn’t need to be convinced. It will enthusiastically build whatever you tell it to build. The PRD is wasted on it, and worse, the narrative format causes real problems. Ambiguity that a human engineer would resolve by asking a question, the agent resolves by guessing. Internal contradictions that a human would flag, the agent just picks one interpretation and runs with it.</p><p>What the agent needs is a declaration, not a set of instructions. If you declare the target customer, a crisp problem statement, a sharp set of ordered user stories with acceptance criteria, and the assumptions that need to be true for this all to work, the agent will surprise you with how creative it can be in the implementation.</p><p>This pattern is already familiar in the software engineering realm: it’s the difference between declarative and imperative. Declarative specs give agents room to be creative within constraints. Imperative specs turn them into a typing service and force them into mistakes because they don’t understand how to avoid them.</p><p>Describe what needs to be true. Let the agent figure out how.</p><h3>Empathy and assumptions: the missing agent context</h3><p>So what goes into a declaration? I went through a long stretch of trial and error before landing on something simple: treat the agent a bit more like a human colleague and answer the fundamental questions they’d need to empathize with the user.</p><p>Who is the customer? What specific problem are we solving for them? What constraints exist? And most importantly: where are the risks? Every feature has hidden assumptions that need to be true for it to succeed. Once you phrase them as testable hypotheses instead of leaving them unsaid in the margins, you’re looping the agent in on the risks. Current models have a smart approach to assumptions; they treat them like requirements. When that data is present from the start, the agent will steer the implementation in a way that makes them all true. If they encounter one they are feeling concerned about, they can even bring it up and help you gather some evidence on the spot.</p><p>Assumptions communicate intent, the things that might be inferred by people on the team, but that agents miss. Agents are basically living in the movie Memento: their medium-term memory is wiped every time they wake up in the morning! Assumptions are the tattoos that keep them oriented.</p><p>The hard part isn’t testing assumptions. Product people know how to do that. The hard part is seeing them in the first place. They’re buried in the solution, invisible until something breaks. Little things like assuming your customer has the technical depth to use an API product, or assuming a marketplace has enough supply to satisfy demand at launch. We don’t see our own bias clearly. We bring optimism when neutrality or pessimism might be wiser.</p><p>AI models are surprisingly good at seeing around these corners. Give them the customer context and a solution, and they’ll pressure test your idea in a pre-mortem style. We leaned on that and built it into the Revelica story map playbook. You’ll almost certainly see at least one assumption you weren’t aware of.</p><p>Once you give the agent a declaration with assumptions called out, it can check its own work against the spec mid-build. It can query for evidence, both for and against an assumption: “what did customers say about this?” and just as importantly, “is there anything that contradicts what we’re betting on?” It can flag when it’s about to make a decision that contradicts a constraint. And it can log what it decided so the next session starts warm instead of cold.</p><h3>What we wanted the whole time</h3><p>This is what speeding up discovery actually looks like. I talk to customers, feed the transcripts into Revelica, and it synthesizes opportunities. I use those to make focused story map specs with the assumptions called out. I compare multiple ideas against each other and kill the ones where the assumptions are too risky. The best decision is a comparison, not an evaluation of a single option in isolation.</p><p>Then I hand the winner to Claude Code. The agent builds with full context. When it hits a decision point, we work together to gather evidence. Discovery and building happen in the same flow instead of in separate phases that don’t talk to each other.</p><p>Teams shipping with agents are moving faster than ever. The question nobody seems to be asking is whether they’re shipping faster toward the right thing, or just producing more waste at higher velocity. If 80% of features were rarely used when shipping was slow, what’s the number now that shipping is fast and discovery is being skipped entirely?</p><p>We made it easier to <a href="https://revelica.com/tools/story-map">try the story map generator now</a>, go ahead and make one without even onboarding! Describe a feature idea, lay out the key customer context and get a set of user stories with assumptions right at the moment in time they could happen to your customer. That’s the context your coding agent is missing. Copy or download it as json, which is possibly the only format I’m aware of that agents like more than markdown.</p><p>If you have feedback, it goes straight to our Slack. I can’t help but drop what I’m doing to read them, even the hot takes.</p><p><em>Originally published at </em><a href="https://revelica.com/blog/coding-agents-have-no-idea-what-customers-want"><em>https://revelica.com</em></a><em> on March 11, 2026.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cdc1ff594602" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Armory Continuous Deployment Self-Hosted and Managed v2.28]]></title>
            <link>https://medium.com/armory/armory-continuous-deployment-self-hosted-and-managed-v2-28-87562d236db?source=rss-d96b66cf1627------2</link>
            <guid isPermaLink="false">https://medium.com/p/87562d236db</guid>
            <category><![CDATA[spinnaker]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[software-deployment]]></category>
            <category><![CDATA[continuous-deployment]]></category>
            <category><![CDATA[terraform]]></category>
            <dc:creator><![CDATA[Randall Abele]]></dc:creator>
            <pubDate>Tue, 02 Aug 2022 11:00:32 GMT</pubDate>
            <atom:updated>2022-08-17T19:39:04.953Z</atom:updated>
            <content:encoded><![CDATA[<h3>Introducing Armory Continuous Deployment Self-Hosted and Managed version 2.28</h3><p>As I contemplate the current economic situation and observe the industry trend of companies tightening their budgets, I’m reminded of the fact that some of the most amazing products come out of teams who were brave enough to build and innovate during challenging times.</p><p>At the core of Armory is the spirit of community and the power that comes from a group of people who share the feeling that bringing new software into the world will solve big problems and enrich the lives of those around us. It is my belief that teams who stay true to their values in moments of adversity are always rewarded even when facing the challenge of choosing a harder path.</p><p><strong>Expanding the Product Line</strong></p><p>At the beginning of 2022, the Armory team became reinspired by the Spinnaker community and its mission to not only focus on Armory’s new declarative offering, Armory Continuous Deployment-as-a-Service, but also to make a significant investment in both Spinnaker and Armory’s enterprise product, which builds on and contributes to the Spinnaker open source project.</p><p>To accommodate the release of Armory Continuous Deployment-as-a-Service into our growing continuous deployment solutions, we have renamed the self-hosted and managed offerings of Armory Enterprise as follows:</p><ul><li>Armory Enterprise becomes <strong>Armory Continuous Deployment Self-Hosted</strong></li><li>Armory Enterprise Managed becomes <strong>Armory Continuous Deployment Managed</strong></li><li>Armory Agent for Kubernetes becomes <strong>Armory Scale Agent for Spinnaker &amp; Kubernetes</strong></li></ul><h3>Introducing Armory CD Self-Hosted 2.28</h3><p>At Armory, the product and engineering teams have a shared instinct to innovate and deliver new features. After speaking with customers and analyzing the market future of Armory CD Self-Hosted, it was obvious that concentrating on the stability and quality of our product was the next best step. It was important to us to invest in the reliability and performance of that fundamental piece of platform technology, which our customers use as a highway to bring delightful software into the world.</p><p>The guiding principle of stability and quality led us to the release of Armory CD Self-Hosted 2.28.0. This long-term support product release encapsulates all the amazing optimizations, features, and fixes from the Spinnaker community, and raises the enterprise-grade quality bar.</p><p>With this release, we have stayed true to our commitment to capturing increasingly complex use cases to ensure solid releases, as we worked on v2.28!</p><p>Here are some highlights:</p><ol><li>Test Automation Infrastructure: We’ve invested heavily in our test automation infrastructure, capturing and covering new complex use cases that emerge from our elite users as integrated platforms evolve. We also added numerous automated unit, integration, CVE, and end-to-end tests that fire on each commit from OSS Spinnaker or the Armory CD Self-Hosted product features.</li><li>Optimized Release Process: The team optimized the release process to allow for incremental and iterative changes (<em>Short Term Support releases</em>) as needed, culminating in twice annual Long Term Support releases that are backed up by a brilliant group of support and solution engineers who know this product like the back of their hand.</li></ol><p>Ultimately, our validation grows alongside the usage of the product. Engineering teams can be confident when they consume 2.28 and subsequent releases that those releases will work with their specific configurations and use cases. This frees engineers to focus on other tasks and feel confident that the platform can deal with everyone’s favorite “growth and scale” problem.</p><h3>Enhancements</h3><h4>Reduce Operational Risk by Automating Policies on Infrastructure Changes with Terraformer Enhancements</h4><p>We heard from several of our Terraformer users that unintended or non-compliant changes to infrastructure hit them where it hurts: the <a href="https://cloud.google.com/blog/products/devops-sre/using-the-four-keys-to-measure-your-devops-performance">DORA metrics</a>. So we brought a solution that allows you to capture Terraform JSON from a `planfile` and act on it in subsequent stages of your pipeline with an ‘Evaluate Variables’ stage. This allows the pipeline to automatically enforce infrastructure policies and test for quality with every infrastructure change so you reduce any operational risk and increase stability. We also added remote backend support for Terraform Enterprise and Terraform Cloud.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/958/0*JUQ2jEyh0bpjEQAJ" /></figure><p>Check out the <a href="https://docs.armory.io/armory-enterprise/spinnaker-user-guides/terraform-use-integration/">Terraformer documentation</a> for details on these new capabilities.</p><h4>Speed Up Your Time-to-Market and Accelerate your Velocity with Parallel Pipelines</h4><p>The second feature we want to highlight is a clever optimization that was driven by one of our customers unique scale use cases, called <strong>Max Concurrent Pipeline Executions</strong>. This was an open-source Spinnaker contribution that makes Armory CD Self-Hosted even better!</p><p>As we worked with our customers pushing the limits of deployments at scale, we saw scenarios where enabling concurrent pipeline executions would cause performance degradations and timeouts, but disabling concurrency was not fast enough for the deployment frequency goals of the team.</p><p>Max Concurrent Pipeline Executions allows you to throttle the number of maximum parallel pipeline executions, so you can dial in your pipeline to maximize deployment frequency without encountering any performance degradations or timeouts. This effectively allows you to continue accelerating your deployment frequency and velocity and speed up your time-to-market with exciting new features and products.</p><p>You can find the Max Concurrent Pipeline Executions feature in your pipeline’s Configuration section. It is disabled by default:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/920/0*g0ehsncXiiiRv2ER" /></figure><h3>What’s next?</h3><ol><li>Check out the <a href="https://docs.armory.io/armory-enterprise/release-notes/rn-armory-spinnaker/armoryspinnaker_v2-28-0/?utm_source=mediumpub&amp;utm_medium=Stories">Armory CD Self-Hosted release notes</a> for all the changes</li><li>Talk to us about making the world better with software in the <a href="https://spinnaker.io/docs/community/#slackhttpjoinspinnakerio">Spinnaker Slack</a>! We’re over in <strong>#armory</strong></li><li>We are so excited about the open source contributions for <em>AWS Launch Templates</em> that we plan to tell that story in a dedicated future post! Stay tuned.</li></ol><p>Thank you for changing the world with cool software. It’s time to build!</p><p>See you out there,</p><p>Randall Abele, Group Product Manager</p><p><em>Originally published at </em><a href="https://www.armory.io/blog/self-hosted-and-managed-v228/?utm_source=mediumpub&amp;utm_medium=Stories"><em>https://www.armory.io </em></a><em>on August 2, 2022.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=87562d236db" width="1" height="1" alt=""><hr><p><a href="https://medium.com/armory/armory-continuous-deployment-self-hosted-and-managed-v2-28-87562d236db">Armory Continuous Deployment Self-Hosted and Managed v2.28</a> was originally published in <a href="https://medium.com/armory">Armory</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>