MST 504Mastery · 30 days3 US credit hours·Mastery course · Upper-division / professional elective·v1 · updated 3 Oct 2026

Jev Mastery — The Decision Layer

University / Board Instructor Student Outcomes & Skills

Course outcomes

CO1

Demonstrate Foundations — decisions, probability, economics competence: complete the Day 1–7 builds, explain the underlying theory, and defend the phase artefact under live questioning.

C2 Understand

Framing decisions vs generation (System One thinking)Probability, calibration, Brier & ECE Choosing Choice / Score / Noul Confidence bands & expected utility Cost modeling & cascades Separating vendor claims from evidence

CO2

Demonstrate Core building — API, SDKs, question design competence: complete the Day 8–15 builds, explain the underlying theory, and defend the phase artefact under live questioning.

C3 Apply

Choosing Choice / Score / Noul Confidence bands & expected utility Raw API use & access paths Python & TypeScript SDKs State design & redaction Question wording iteration Speculative fan-out & composite scoring Version pinning, rate limits, fallback

CO3

Demonstrate Integration & graphs — agents, knowledge graphs, evals competence: complete the Day 16–23 builds, explain the underlying theory, and defend the phase artefact under live questioning.

C5 Evaluate

Probability, calibration, Brier & ECE Cost modeling & cascades Model selection, routing, fallback and provider portability LangChain & LangGraph integration Knowledge-graph traversal (Neo4j)RAG pipeline design Gateways & ecosystem selection LLM-as-judge calibration

CO4

Demonstrate Production, AWS, security & mastery competence: complete the Day 24–30 builds, explain the underlying theory, and defend the phase artefact under live questioning.

C6 Create

Cost modeling & cascades Separating vendor claims from evidence Version pinning, rate limits, fallback Reference architecture & trust boundaries AWS deployment with Bedrock fallback AI red-teaming Vendor risk & data governance Portability & open alternatives Teaching & continuous evaluation

Phases & skills

Phase	Topic	Outcomes	Deliverable	Skills
1	Foundations — decisions, probability, economics Days 1–7 · C2	CO1	CAPSTONE (Phase 1) — Decision brief with honest limits	Framing decisions vs generation (System One thinking)Probability, calibration, Brier & ECE Choosing Choice / Score / Noul Confidence bands & expected utility Cost modeling & cascades Separating vendor claims from evidence
2	Core building — API, SDKs, question design Days 8–15 · C3	CO2	CAPSTONE (Phase 2) — Support-triage decision service	Choosing Choice / Score / Noul Confidence bands & expected utility Raw API use & access paths Python & TypeScript SDKs State design & redaction Question wording iteration Speculative fan-out & composite scoring Version pinning, rate limits, fallback
3	Integration & graphs — agents, knowledge graphs, evals Days 16–23 · C5	CO3	CAPSTONE (Phase 3) — Graph agent with Jev at every edge	Probability, calibration, Brier & ECE Cost modeling & cascades Model selection, routing, fallback and provider portability LangChain & LangGraph integration Knowledge-graph traversal (Neo4j)RAG pipeline design Gateways & ecosystem selection LLM-as-judge calibration
4	Production, AWS, security & mastery Days 24–30 · C6	CO4	MASTERY CAPSTONE — Your track's final build	Cost modeling & cascades Separating vendor claims from evidence Version pinning, rate limits, fallback Reference architecture & trust boundaries AWS deployment with Bedrock fallback AI red-teaming Vendor risk & data governance Portability & open alternatives Teaching & continuous evaluation