Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by Theo Valmis
When Architecture Decisions Actually Apply
In my previous Radar article, “Architectural Guardrails for AI-Generated Code,” I argued that AI coding agents need a layer that can hold architectural decisions in a structured corpus, retrieve the relevant ones, put them into the agent’s context before it writes code, and enforce explicit constraints with reproducible checks. Testing that model exposed two distinctions I had compressed together.
Architectural Guardrails for AI-Generated Code
Consider a composite of a failure pattern that’s becoming increasingly common on teams that have scaled AI-assisted development past a handful of enthusiasts. A staff engineer named Priya opens a pull request. The PR is 340 lines and adds an endpoint that writes to the customer table directly, bypassing the internal customer service API. The code is clean. The tests pass. The AI coding agent that wrote it has been the team’s most productive contributor for six months.
Architectural Guardrails for AI-Generated Code
2 hours ago CHICAGO (AP) — U.S. President Donald Trump has asked Apple to change the name of Lake Ontario to “Lake America” on its maps app, his administration’s interior chief said Monday, days after signing an executive order over the body of water now in the spotlight amid his escalating trade war with …
Anthropic's AI-Native SDLC Has Three Controls. It's Missing a Fourth.
Anthropic published an AI-Native SDLC playbook last week. Its central claim: code generation is no longer the slow part of delivery. The stages around it, planning, review, security sign-off, release, still run at human speed. Once agents write most of the diff, that mismatch has to break one of two ways: review queues grow until the speed gains disappear, or code ships under-reviewed. Neither is acceptable if you're regulated, and neither is great even if you're not.
How Mneme governs AI-generated code before the model writes a line
LLMs start every call from zero. They reintroduce a library you dropped six months ago, rebuild a component you chose to keep small, and contradict decisions your team already settled. Each violation reads as reasonable on its own. Stack them across a week of agent sessions and you get architectural drift. Mneme works at the prompt boundary. It reads the decisions your project already made and checks the task against them before the model generates anything.
How to Turn a Recipe Screenshot into Structured JSON
Anyone building a cooking, meal planning, or grocery app hits the same wall fast: people do not save recipes as clean data. They save them as pictures. A screenshot of a Pinterest pin. A frame grabbed from a TikTok. A photo of a page in a cookbook. A wall of text pasted into a note. If your app needs ingredients, servings, and steps as real fields, that messy input is a genuine problem. Scraping a web page only works when there is a web page, and most recipes shared on social apps never have one.
Are We Going to an AI-First Frameworks Era?
For a long time, web development has been a constant battle with syntax. From writing verbose CSS files to memorizing the endless classes of libraries like Tailwind, developers have remained tethered to the "machine’s way" of speaking rather than expressing their own creative vision. However, with the massive leaps in Large Language Models (LLMs), we are standing at the threshold of a completely new era: the AI-First Framework era.
Does the EU AI Act apply to my chatbot?
I originally published this on disclosekit.dev — I'm building tooling for this deadline and the research surprised me enough to share. June 2026 · 5 min read Short answer: if your chatbot talks to people in the EU, probably yes. From August 2, 2026, Article 50(1) of the AI Act requires you to tell users they are interacting with an AI system. Not in your terms of service. At the start of the conversation, where they can actually see it.
AI in Software Engineering: Why Understanding Still Matters
The Limits of AI in Software Engineering: Why Understanding Still Matters A software company delivers a product to its customers. Then a new government regulation appears, mandating changes to that software by a specific deadline. The company creates backlog items, breaks them into tasks, and developers start coding. The question is: how much AI should they use to write that code? Three Levels of AI Adoption Level one.
Claude Fable 5: What It Is and What It Means for Developers
Anthropic launched Claude Fable 5 today, June 9, 2026. It is the most capable model Anthropic has ever made generally available, and it lands at a price point below half of what Claude Mythos Preview costs. If you build on Claude, or if you are evaluating which frontier model to put at the center of your stack, this release changes the calculus. Here is everything you need to know, sourced directly from Anthropic's announcement. What Is Claude Fable 5?
The Junior Dev Who Never Had to Google Anything — Is That a Superpower or a Problem?
Let me tell you four things that happened to me. One. I sat in a mock interview. The question was basic — embarrassingly basic for someone who'd been coding for months. And I went completely blank. Not because I didn't know the concept. I did. I'd used it dozens of times. But every single time I'd used it, I'd asked AI first. I never had to hold the answer in my own head. So when someone asked me to pull it out of my head on the spot, there was nothing there. Two.
Cursor Developer Habits Report 2026: Why AI Coding Needs Governance Infrastructure
Cursor's Developer Habits Report is one of the clearest signals yet that AI coding has crossed from individual productivity into software-delivery infrastructure. The headline numbers read as a story about speed: more code per week, larger PRs, deeper agent sessions, more changes committing without manual review. The deeper implication is governance -- whether teams can preserve architectural intent while generation, review, automation, and commit flows all accelerate at once.
Agent Runtime Governance: The Next AI Infrastructure Layer
Google's Managed Agents announcement is one of the clearest signals yet that the AI industry is moving beyond stateless tool calling toward persistent execution environments and long-running agent systems. That shift expands what models can do. It also expands the governance surface -- from prompt and PR review into the runtime itself. We spent two years building brains in jars For most of the current AI cycle, the system around the model has been thin.
Microsoft's Agentic Transformation Playbook Shows Why AI Agent Governance Is Now Infrastructure
Microsoft's Agentic Transformation Patterns Playbook is a useful signal because it does not treat AI agents as another productivity tool. It frames agentic AI as an enterprise operating-model shift: agents are moving from assisting humans to executing work across processes, systems, and teams.
The Acceleration Whiplash and the Governance Gap
The Faros AI Engineering Report 2026 is not a survey of developer sentiment. It is two years of telemetry from 22,000 developers across 4,000 teams, measuring what AI adoption actually produces downstream. The findings have a name: the Acceleration Whiplash. The structural explanation has one too. What the telemetry actually shows The output numbers in the Faros report are real and worth stating plainly. Epics completed per developer are up 66.2%. Task throughput per developer is up 33.7%.
Executable Architectural Intent: The Promotion Path From Docs to Constraints
Most project knowledge wants to be findable. A smaller, more important subset has to be binding. Executable architectural intent is the name for that subset — the slice of architectural knowledge that has been promoted out of documentation and into the layer where it can actually constrain what an AI agent does. Teams writing for AI coding agents accumulate project knowledge fast: ADRs, PRDs, design notes, AGENTS.md files, NotebookLM corpora, Cursor rules, repo-native wikis.
Why I Built Mneme HQ: Preventing AI Agent Architectural Drift
Originally published on theovalmis.com. Every time you start a new session with an AI coding agent, it has forgotten everything. Not just the small things — the names, the syntax, the last error message. It has forgotten the decisions you made three weeks ago about why you chose Postgres over MongoDB. It has forgotten the auth pattern you locked in after a security review. It has forgotten that you explicitly ruled out a particular library because it caused problems in production.
Start Here: Why AI-Assisted Development Needs a Governance Layer
Every major shift in software engineering produces a new infrastructure requirement. The cloud era gave us orchestration. The DevOps era gave us CI/CD. The generative AI era is producing its own requirement — and most teams haven't named it yet. That unnamed requirement is governance. Not memory. Not context injection. Not longer system prompts.
Why Code Review Cannot Scale With AI Output
AI coding assistants have made generation cheap. They haven't made review cheap. The result is a compounding bottleneck that most engineering teams are only beginning to feel — and that no amount of hiring will resolve. For most of software engineering history, writing code was the bottleneck. Senior engineers reviewed what juniors wrote, and the review burden was proportional to human writing speed. The ratio was manageable. AI coding assistants break this assumption.
Review Is Not Governance
CodeRabbit helps review AI-generated code. Mneme helps govern what the AI generates in the first place. These are not competing tools. They are different layers of the same problem. Most teams using AI coding assistants have invested in the review layer. Tools like CodeRabbit, PR comment bots, and AI-assisted linters sit at the end of the pipeline, inspecting output after it surfaces in a pull request. That is useful work. But it is reactive by design.
Harness Engineering Still Needs Governance
The industry has moved from prompt engineering to harness engineering: execution systems that coordinate models, tools, memory, and retries across long-running agent loops. Harnesses solve how agents act. They do not solve whether the actions stay within architectural boundaries. As autonomous workflows scale, the missing layer is governance infrastructure. The shift: prompts, harnesses, execution systems Three years ago, the prevailing question was how to phrase a prompt.
Long-running agents need more than memory
Anthropic's managed-agent harness solves one hard problem: continuity. Progress logs, feature lists, git checkpoints, and startup scripts give each new session a map of what happened. But continuity is not governance.
Code Got Cheaper. Our Rituals Didn’t Notice.
Much of the modern software development lifecycle was built around code being expensive and hard to change. So we built ceremonies, rituals, and whole philosophies around deciding what to build before anyone built it. We planned every story long before an engineer started writing any code. Product managers would curate an extensive backlog of work just waiting for engineering resources. Designers would build mockups because doing so was so much quicker than implementation.
Datadog's State of AI Engineering Report Quietly Confirms the Governance Crisis
Datadog surveyed over 1,000 organizations running AI in production. The report is framed around observability and operational maturity. Read carefully, it is also the clearest empirical signal yet that the industry's next unsolved problem is governance. Most industry reports on AI engineering measure what is easy to measure: adoption rates, token volumes, model preferences, framework usage.
Why AI Architectural Governance Needs Precedence Semantics
Two architectural decisions overlap. An engineer in Cursor follows one. The async PR bot in CI follows the other. A reviewer signs off on a diff that flatly contradicts a decision that landed last week. Nobody is wrong. Nobody has the authority to be right. This is what AI coding governance looks like without a precedence layer — and it is the gap every system in the category is currently failing to close. Most of the conversation about "AI coding governance" in 2026 is still about the wrong layer.
Memory Is Not Governance
The AI coding category has spent two years calling four different systems by the same word. Memory. Context. Retrieval. Governance. They share some primitives. They optimize for different things. And the most expensive mistake an engineering team can currently make is buying a memory system and expecting it to govern. The AI coding category is awash in memory products. Letta. Mem0. OpenAI's memory feature. Cursor's per-user context. Claude's projects.
Why RAG Fails for Architectural Governance
Retrieval-augmented generation is an excellent tool for knowledge lookup. It is the wrong tool for enforcing architectural decisions. The distinction matters — and most teams building AI coding workflows haven't confronted it yet. When teams first encounter the problem of governing AI-generated code, RAG is the intuitive answer. You have a set of architectural decisions — ADRs, style guides, internal wikis, team conventions — and you want your AI coding assistant to respect them.
Autonomous Code Remediation Requires Architectural Governance
For the last two years, the central question in software engineering has been: can AI generate production-quality code? The answer is increasingly yes. But that answer unlocks a harder problem -- one the industry is not yet organized around: How do autonomous remediation loops stay architecturally stable? The Shift Has Already Happened The software lifecycle is already partially autonomous. Not in a speculative sense. In a visible, operational sense.
I shipped an AI-generated PR that violated four of our own architecture decisions. Nobody caught it
`--- title: "I shipped an AI-generated PR that violated four of our own architecture decisions. Nobody caught it." published: false description: "AI coding agents don't make architectural mistakes. They make architecturally unconstrained outputs — and PR review was never sized for the throughput." tags: ai, architecture, devops, productivity The PR was 812 lines. Eleven files. Tests green.
What Cannabis Retail Pricing Is Telling Us About Market Pressure in 2026
In the cannabis industry, pricing behavior often reveals more about market pressure than financial statements ever could. However, some of the most revealing cost signals are subtler, showing up in pricing behavior, discount strategies, and category-level decisions rather than on balance sheets. The CannabisDeals US Cannabis Price Index, launched in December 2025, tracks these signals in real time by monitoring more than 21,000 products across 57 licensed online retailers.
Show More
loading
Actions
Is this you?
As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.Get in touch with Theo
Contact Theo, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack