Writing

Long-form essays on agent architecture, calibration, and the patterns that hold up in production.


Calibration, boundary layers, and the distance between yes and no: a practitioner’s geometry of LLM reliability

The geometric framing that ties context decay, retrieval failure, memory architecture, and agent calibration into one system. · 12 min · 2026-04-26

Two-tier memory for production agents: what chat systems don’t tell you

An architectural pattern for agents that need to remember across sessions without choking on their own context. · 13 min · 2026-04-26

The 6-tier agent maturity model: why most enterprise AI is failing on a calibration mismatch, not a tooling gap

A practitioner’s tier ladder for self-locating an AI initiative — and the diagnostic that names the canonical enterprise failure in three questions. · 18 min · 2026-04-26

More essays

Attention management: routing between native LLM capability and augmented skills and tools

What a production agent system decides at every step — when to trust the model, when to reach for a skill, when to call a tool. · 9 min · 2026-04-26

The context window is a battery: surviving compaction on long-running agents

What happens to your calibration when the context fills up, why compaction drops the work you need most, and the operational discipline that protects you from it. · 7 min · 2026-04-26