Five-week Certified AI Engineering cohort for senior engineers to begin July 25, 2026

A five-week Certified AI Engineering cohort designed for senior engineers and technical leaders working on production AI systems will begin July 25, 2026, running four hours per week in live online sessions. Enrollment is open now; weekly meetings are scheduled on Saturdays at 9:00 AM PDT. The program aims to help senior contributors confront concrete production trade — offs—from retrieval and inference routing to reliability and ownership — within a structured, practitioner — focused setting.

The offering pairs a certification outcome with an intentionally small, confidential cohort format. Sessions are structured as peer groups so participants can discuss real decisions from their own organizations rather than inspect generic case studies, and the curriculum emphasizes applied frameworks that participants can use immediately on platform and product problems. Hien Luu, senior engineering manager at Zoox and author of MLOps with Ray and Beginning Apache Spark 3, will facilitate the cohort. Luu leads Zoox’s Machine Learning Platform team and has presented on AI/ML infrastructure at QCon; his facilitator role centers on steering peer critique of architectural and organizational trade — offs and guiding application of proven frameworks to participants’ production challenges.

The curriculum is organized by week. Week 1 covers becoming an AI-native engineering team, including shifts in product and architectural habits. Week 2 focuses on designing retrieval — augmented generation (RAG) and context pipelines, addressing retrieval architecture, knowledge graphs, and memory pipelines. Week 3 examines single — purpose agents and multi — agent orchestration. Week 4 takes up AI platform and infrastructure decisions — centralization versus federation and workload routing. Week 5 concludes with operational excellence topics: evaluation, reliability, and ownership.

The program responds to an industry gap where teams moving features from prototype to production often make crucial choices — about retrieval architecture, context pipelines, agent orchestration, evaluation, inference cost, observability, and operational ownership — without external benchmarking or peer scrutiny. For senior practitioners, the concrete value is applying frameworks to live problems and receiving structured, confidential feedback on architectural decisions and technical debt. either new platform and reliability approaches or confirmation that existing trade — offs are appropriate, plus clearer guidance on cost, inference routing, observability, and operational responsibilities.