Writing

Notes on getting AI into production.

Two threads run through most of these: why AI installs fail for organizational reasons, and how the production controls that fix them are actually built. No hype, working detail.

• Thesis6 min

  The alignment problem most companies have is organizational

  Model alignment is the labs' job. The version that breaks production is an organization that never decided what its AI was for.

• Thesis5 min

  Why every AI sales call leaves the room more confused

  The vocabulary grows every quarter while understanding shrinks. For a CEO or CTO trying to underwrite a decision, the bottleneck is not capability. It is clarity.

• Technical9 min

  Every AI platform converges on a control plane. Build it on purpose.

  Multi-team LLM traffic creates nine predictable problems that all want to live in one place. Whether that place is a single gateway, several gateways feeding a shared plane, or a runtime layer is an open question. The decision to design it is not.

• Technical10 min

  Centralized guardrails across three clouds, and the cost of a new vendor

  Each cloud's native safety tool solves a different slice. Putting guardrails in one place means a separate product, and that is a procurement and trust decision as much as a technical one.

• Technical9 min

  When nobody reviews the output: evaluating agents in production

  With a coding assistant a developer checks every change. With an autonomous agent, no one does, so the eval suite becomes the quality gate: capability and regression suites, code, model, and human graders, outcomes over transcripts.