Writing

Notes on ML systems, mostly drawn from production work.

• Expensive ways small ML teams imitate big tech

  10 May 2026

  Why small ML teams burn cycles on practices calibrated to big-tech scale, and what to do instead.