Source, Sink, Drain: an IBM engineer's blueprint for running an AI software factory
TejasKumar_ · x · 2026-11-13
- IBM AI engineer Tejas Kumar distills his AI software factory into three parts: source (you decide what gets built in GENESIS.md), sink (a pm agent turns it into a Linear issue queue), drain (parallel coding agents burn through tickets as fast as token quota allows).
- Running this on Daiku, his startup giving AI its own browser, grew the board from 475 tickets on Oct 16 to 921.
- Vendor definitions from Cortex and LaunchDarkly name the concept but miss the shape you need when wiring it up at 11pm; the framework also answers who's left in the factory once agents do the rest.
More from coding & agent
- Open-source flight search MCP server refuses to answer what it can't confirm, curbing model hallucination — Fast_Entrance7839 · 2026-11-13
- Amplitude launches Wave: AI that decides what to build next from product data — testingcatalog · 2026-11-13
- jxnlco: when AI makes you faster, go deep on new things and grow your taste — jxnlco · 2026-11-13
- Vibe-coded space-defense game ORBITAL SIEGE shows Three.js gameplay from one prompt — kimmonismus · 2026-11-13
- AI-generated WebGL2 shader bends light around a black hole on the first try — kimmonismus · 2026-11-13
- Sententos puts a personal agent in your Mac's notch that can take over your screen — ycombinator · 2026-11-13