ESSAYS
Technical essays on AI systems that work outside demos.
Long-form writing on agents in production, workflow design, model/tool tradeoffs, and the engineering behind trustworthy AI products.
Essay
Why most AI products break after the demo
Technical essays on reliability, retries, orchestration, observability, and trust in LLM systems.
Series
Agents in production
Architecture notes on state, permissions, memory, tool use, and the ugly realities of execution.
Operator note
Workflow design around LLMs
The system layer around prompts: contracts, queues, verifiers, retries, and human override.