DO the work. Don't just talk about the work
Four questions I run on my own work before I run them on anybody else's. Including the one that made me rewrite an issue hours before it shipped.
Signal Over Noise — by Doneyli
Practitioner AI insights from a Principal AI Architect with 20+ years in data and AI. Free: frameworks and signal analysis. Paid: full "I Built X" tutorials with production code. Less Noise. More Signal.
Live Last read · last published · next check
Latest posts
Are You Improving the Judge or the Agent?
Most teams don't have an evaluator shortage. They have an evaluation-system design problem, and it starts with which of two targets you're aiming at.
My LLM Judge Lied to me. Twice.
A judge missing context does not return unsure. It returns 0.85 and a fluent paragraph explaining why.
Sovereign AI Starts With Your Traces, Not Your GPUs
5:21 pm Eastern, June 12. Anthropic receives a directive from the US government. Within hours, two of its frontier models, Fable 5 and Mythos 5, go dark for every customer on the planet. Not a rate limit. Not a regional outage. Off. Subscribe now I read the status page from my kitchen table and had the same thought a lot of you probably had: I use one of those models every day. And then a quieter…
How to Certify AI Agents
and WHY they don't stay certified
Agents. Agents. Agents.
Field notes from three days with 7,000 AI engineers in San Francisco: what’s real, what’s bubble, and the five signals worth your attention.
Get Discovered by AI. Steal the Playbook.
I asked Claude to recommend a tool in a category I had spent six weeks building a product for. It named three companies. The one I had built was not one of them, even though the homepage said exactly what the buyer was searching for, in plain English. The product was not the problem. The problem was that the agent doing the searching had never been allowed to read the site, and the crawlers that…
Your AI Agents Have a Governance Problem
The three failure modes killing enterprise AI projects before they ship
The Model Doesn't Matter Anymore
The harness is the part you actually own. Here are the four layers.
The AI Migration That Documented Itself
Moving a Docker Home Server from MacBook to Mac Studio