← Back to home/nextWhere I'm headed3 steps, starting now — the plan behind these notes.Open →
// Blog
Writing on engineering.
Practical notes from building production systems — AI agents, distributed backends, and the design decisions that hold up under load.
Index4 of 4 entries
Nov 10, 2026Building a ReAct-Style Agent Loop with Tool Calling3 minHow Northstar runs natural-language board operations through a server-side agent loop — and why the model's output is treated as untrusted input.AI EngineeringAgentic SystemsNorthstarNov 8, 2026Server-Authoritative State: Treating AI Output as Untrusted Input2 minAn AI agent that can move cards on a Kanban board is only useful if it cannot corrupt the board. Designing the state machine that stands between the model and the data.AI EngineeringArchitectureState MachinesNov 6, 2026Streaming LLM Responses with Server-Sent Events2 minWhy SSE beat WebSockets for Northstar's agent activity feed, and the trade-offs of streaming partial tool calls over a unidirectional channel.AI EngineeringBackendSSENov 4, 2026What the AI Engineer for Developers Associate Actually Covers2 minA breakdown of the DataCamp associate certification — the ten courses, the 29 hours, and the parts that changed how I build LLM features.AI EngineeringCertificationCareer