The Human in the Loop
What happens when a platform engineer hands the pager to agents — experiments, war stories, and the occasional human override.
The story so far: from chatbots to agents
Every few months a new agent framework lands, and the same question fills my feed: which one should I pick? LangChain, LangGraph, CrewAI, ADK, OpenAI's SDK — the names rotate, the anxiety doesn't.
Nine of Ten Steps: Why Agent Reliability Is an Evals Problem
It's 2 AM. Production is down. Error rates are spiking across regions, transactions are failing, and the incident has auto-escalated. A SWAT bridge spins up — fifty people on the call: SREs…
The Hundred-Year Conversation: A Century of Speaking Machine — Until Machines Spoke Human
For a hundred years, using a computer meant learning its language — levers, punch cards, cursors, menus, apps. Then, in 2022, the direction of that conversation reversed. This is the story of how we…
Autoresearch: How Deep-Research AI Agents Are Actually Built
Everyone is shipping "deep research" agents, but the architectures underneath are surprisingly different. This post maps the autoresearch pattern, the state of the art, the benchmarks, the critiques…
Evolution of Prompt Engineering to Context Engineering
Interacting with large language models used to be a craft of wording. Now, with million-token windows and AI assistants living inside Outlook, Teams and VS Code, the game has quietly moved from…