What stakeholder demos caught that our tests couldn't
Four kinds of failure our test suite couldn't see, which only surfaced once real stakeholders used an AI booking assistant in a room.
AI
13 notes filed here
Designing a logging system for an AI service
Building a standardised logging system for an AI assistant service: making consistent structured logging the path of least resistance, and thinking carefully about what an AI service should and shouldn't persist.
From merged pull requests to a performance review
Merged pull requests already record what you did. Turning them into a work log with an LLM is easy; the hard part is the defensive rules that stop it corrupting the vault.
Three bugs hiding in a cross-service feature
Building a feature across four services meant coordinating an AI backend, a fare search service, and two frontend components. Along the way: a silent configuration bug, a duplicate event contract, and an HTML injection path.
Folding a conversational search selection into a basket
Saving an option from a conversational flight search into a booking basket needed a new route: a partial update that has to merge a selection into whatever is already there. Plus privacy-aware MCP observability, Copilot context hooks and a naming cleanup.
Why our AI assistant's orchestration lives in code
A reflection on where coordination logic should live when building AI assistants into production systems, and why application-layer orchestration beats prompt-driven coordination.