Mastery track of the AI series. The capstone of the v1 curriculum — adds the production primitives that separate a demo from a deployable script: embeddings (cosine similarity, pairwise comparisons), chunking and storing vectors, semantic search, full RAG pipelines with citations, RAG failure modes, eval suites at scale, model routing (cheap vs strong), caching by prompt hash, parallel LLM calls via `concurrent.futures`, PII detection and redaction, rate-limit-aware retry, fallback chains, human-in-the-loop approval gates, multi-step agents with recovery, and a synthesis that composes 8+ primitives across the full Python + Automation + AI stack. 30 lessons across 4 weeks. Requires AI Patterns. Max tier (Sandbox + OpenRouter + Composio).
Week 1
Week 2
Week 3
Week 4