Notes from the engine room.
Essays and real engagements. How platform foundations get built, where velocity actually comes from, and what survives when systems hit their first 10× moment.
Two Tools, Any Question: Building Axioma's MCP on a Semantic Layer
Why we skipped both text-to-SQL and one-tool-per-chart for Axioma's MCP, built two general tools on a Cube semantic layer instead, and let the schema descriptions do the real work.
Two Tools, Any Question: Building Axioma's MCP on a Semantic Layer
Why we skipped both text-to-SQL and one-tool-per-chart for Axioma's MCP, built two general tools on a Cube semantic layer instead, and let the schema descriptions do the real work.
How we built a video ingestion pipeline for 60× scale headroom
Building a high-volume video ingestion system for a scale target that refused to hold still, and the three times it pushed back and taught us something.
Your AI Agent Should Not Grade Itself: Software Self-Verification Crisis
I built Boundary-First Engineering around one annoying question: who verifies the code when AI wrote the checks too? My answer is boundaries, contracts, and evidence that came from outside the implementation loop.