Newsletter
What you’ll get
Topics covered:
- Production ML systems — Control planes, rollback patterns, observability-by-default
- GenAI evaluation & safety — LLM-as-judge, evaluation loops, safety rails
- Causal measurement — Incrementality, geo experiments, lift estimation
- MLOps patterns — CI/CD/CT for models, contract-driven development
- Platform thinking — Architecture patterns that scale teams
Frequency: Published when there’s something worth sharing (typically 1-2 posts per month).
No spam: Only substantive technical posts. No promotional content, no link dumps.
Recent posts
Operating GenAI safety and policy reviews
April 20, 2025
GenAI systems drift as prompts, tools, and models change. Safety operations keep that drift controlled without slowing teams down.
Evaluation blueprints for GenAI systems
April 06, 2025
GenAI features fail quietly unless evaluation is baked into delivery. A good blueprint pairs offline evals (checklists, red-team prompts, golden questions) with online signals (satisfaction, refusals, latency, cost) and makes...
Backtesting ML pipelines before rollout
March 18, 2025
Backtesting bridges the gap between offline metrics and production behavior. It prevents surprises by replaying real workloads through new code and models.
Continue the conversation
Need a sounding board for ML, GenAI, or measurement decisions? Reach out or follow along with new playbooks.