#full-stack

I Built a Full-Stack LLM Observability Platform for a Founding Engineer Take-Home

Every inference logged with zero hot-path overhead — SSE streaming, fire-and-forget ingest, and hourly rollups in PostgreSQL.