Skip to content

Curriculum — Master Case Catalog

HungerSync Curriculum — Master Case Catalog (v4, all four certs folded in)

Section titled “HungerSync Curriculum — Master Case Catalog (v4, all four certs folded in)”

Single source of truth for the case series. Supersedes the lineups in case-map.md and coverage-v2/v3. Those remain as rationale; this is the authoritative catalog.

Operating principles

  • World-neutral cases. Every case brief below is a business/operational problem in plain domain language — no vendor, cloud, or service names. Cert coverage is an overlay (the tag columns), never baked into the problem.
  • Everything optional, with visibility. You can work any case or any layer independently. The coverage dashboard shows exactly what each case teaches so nothing is a black box.
  • Four interlocking certs, four layers. Data platform → Prediction → Application → Agent.

CertificationTask statementsCovered by seriesNotes
Claude Certified Architect – Foundations3030 / 30 (100%)full
AWS AIP-C01 (GenAI Developer Pro)2020 / 20 (100%)full
AWS MLA-C01 (ML Engineer Assoc.)1212 / 12 (100%)full
AWS DEA-C01 (Data Engineer Assoc.)179 directly + 8 absorbed (~70% by weight)subset — data-platform core full; D4 governance & some D1/D3 incidental

Per-task indices in §4. DEA is intentionally a subset (see coverage-v3).


LayerCertOwnsCases
0 · Data platformDEA (subset)the lakehouse: ingest→store→catalog→model→serveDE1, DE2
1 · PredictionMLA-C01delay / demand / taste modelsML1, ML2, ML3
2 · ApplicationAIP-C01GenAI app: discovery, RAG, safety, ops, integrationCS1, CS4, CS8
3 · AgentClaude Architectresolution agent, dispatch orchestration, dev toolingCS2, CS5, CS6, CS7
spinejointcross-layer incidentsCS3, CS9
capstoneall fourend-to-endCS10

Reads upward: lakehouse feeds models, models feed app, app drives agent.


3. Case catalog (neutral briefs + coverage overlay)

Section titled “3. Case catalog (neutral briefs + coverage overlay)”

C=Claude · A=AIP · M=MLA · D=DEA. “Tasks” lists covered task-statement IDs per cert.

DE1 · The lakehouse: storing & modeling the data Brief: HungerSync must land flight, weather, vendor, order, and voucher data from many feeds into one durable, queryable foundation, then model it so analysts, data scientists, and the app all read the same facts. Decisions: where each kind of data lives, how it’s cataloged and discovered, how it’s dimensionally modeled (conformed dimensions — flight, gate, vendor, passenger-segment, time; facts — orders, deliveries, vouchers, delays), lifecycle/retention, and data-quality gates. Tasks: D 1.1, 2.1, 2.2, 2.3, 2.4, 3.4 · absorbs M 1.1 (platform side). Exhibit: capability-map. Assignments: 2.

DE2 · Pipelines & analytics: orchestrating and serving Brief: the foundation must refresh reliably and answer business questions — revenue by concourse, demand by terminal, voucher redemption rates. Decisions: pipeline orchestration and fault tolerance, scheduling, and the analytics/serving surface. Tasks: D 1.3, 3.1, 3.2. Exhibit: value-stream. Assignments: 1–2.

ML1 · Will this flight be late? Brief: predict disruption and gate-dwell from the lakehouse plus live edge feeds; engineer the features; choose a modeling approach. Consumes DE1 (does not re-ingest). Tasks: M 1.1, 1.2, 1.3, 2.1. Exhibit: value-stream. Assignments: 2.

ML2 · Training, tuning, and trusting the model Brief: train and tune the delay/demand models and the menu/cuisine text classifier; evaluate honestly (precision/recall on “late,” forecast error, bias across routes and passenger segments); fine-tune a small model for messy vendor-menu text. Tasks: M 2.2, 2.3. Exhibit: swot (bias/limitations). Assignments: 2.

ML3 · Shipping and watching the model Brief: serve predictions (real-time for live ordering, batch for the nightly forecast) and retrain when drift appears (weather-regime shifts, schedule changes). Tasks: M 3.1, 3.3, 4.1 · A 2.2 (deploy). Exhibit: value-stream. Assignments: 1–2.

CS1 · Designing HungerSync (opener) Brief: turn the funded business case into an architecture under a well-architected lens before the contract clock starts — proof-of-concept, approach selection, deployment and integration strategy. Tasks: A 1.1, 2.2, 2.3 · M 3.2 (IaC) · C (arch judgment). Exhibit: BMC + value-chain. Assignments: 1.

CS4 · Discovery you can trust Brief: the discovery feed must recommend only available items and surface what a flight actually craves — stale-availability and retrieval-quality failures. Tasks: A 1.3, 1.4, 1.5, 3.1.3, 4.2, 5.1, 5.2 · C 2.4, 4.1, 4.2, 4.3, 4.4. Exhibit: journey. Assignments: 2.

CS8 · Trust & Safety: guardrails, privacy, governance Brief: prevent abuse and allergen harm, hold the aggregate-first privacy boundary, and satisfy auditors on who-saw-what and data residency. Tasks: A 3.1, 3.2, 3.3, 3.4, 2.3.3 · M 1.3, 4.3 · D 4.1–4.5 · C 4.1, 5.2. Exhibit: capability-map. Assignments: 2.

CS2 · The ordering agent you can trust Brief: the resolution agent applies a voucher to the wrong passenger and confuses two similar lookups; reliability via enforcement gates and tool clarity. Tasks: C 1.1, 1.4, 1.5, 2.1, 2.3, 4.2 · A 2.1, 2.3, 2.4. Exhibit: ecosystem + revenue-flow. Assignments: 1.

CS5 · The four-hour delay conversation Brief: a long conversation drifts and loses a dietary constraint stated in hour one; context preservation, escalation, and resumption. Tasks: C 1.7, 5.1, 5.2 · A 1.6, 4.1, 5.2. Exhibit: journey. Assignments: 1.

CS6 · Claude Code: config & workflows Brief: the platform team’s conventions must not drift across the monorepo; plan vs direct execution for a big dispatch-engine refactor. Tasks: C 2.5, 3.1, 3.2, 3.3, 3.4, 3.5, 5.4 · M/D IaC tag. Exhibit: (code). Assignments: 1.

CS7 · Claude in the pipeline: review & CI Brief: settlement code reviewed automatically in CI without false-positive noise or self-review blind spots. Tasks: C 3.6, 4.6 · A 2.3.5, 2.5.4, 2.5.6 · M 3.3, D 1.4 (CI tags). Exhibit: (code). Assignments: 1.

CS3 · The storm breaks the pipeline Brief: during a ground stop the multi-agent prediction/dispatch pipeline both mis-decomposes coverage and loses a data feed; task decomposition, error propagation, graceful degradation. Tasks: C 1.2, 1.3, 1.6, 2.2, 5.3, 5.6 · A 1.2, 1.3, 2.1, 2.4, 2.5. Exhibit: value-stream. Assignments: 2.

CS9 · The ops center Brief: token spend spikes during the storm, a hallucination reaches a passenger, and the nightly forecast must not block the live floor; cost, monitoring/drift, evaluation, human-on-the-loop review. Tasks: C 4.5, 4.6, 5.5 · A 2.2, 4.1, 4.2, 4.3, 5.1 · M 4.2 · D 3.3. Exhibit: swot. Assignments: 2.

CS10 · The next ground stop (end-to-end)CONFIRMED: spans all four layers. Brief: a fresh major IROPS event exercises the whole vertical at once, and the system holds. Beats by layer:

  • L0 Data platform (D): the lakehouse serves fresh, conformed facts under load; late-arriving feeds reconciled; analytics still answer “where is demand forming.”
  • L1 Prediction (M): models predict delay/dwell/demand and emit a pre-positioning plan; drift watch holds as the weather regime shifts mid-event.
  • L2 Application (A): discovery stays grounded (no 86’d items), guardrails hold, the airline voucher rail fires, ops/cost stay bounded.
  • L3 Agent (C): the resolution agent handles surge conversations and the dispatch coordinator orchestrates fulfillment with human-on-the-loop remote ops. Tasks: integrative across C / A / M / D. Exhibit: all. Assignments: 1 capstone project.

4. Master task → case index (full visibility)

Section titled “4. Master task → case index (full visibility)”

Claude (30/30): 1.1 CS2 · 1.2 CS3 · 1.3 CS3 · 1.4 CS2 · 1.5 CS2 · 1.6 CS3 · 1.7 CS5 · 2.1 CS2 · 2.2 CS3 · 2.3 CS2 · 2.4 CS4/CS6 · 2.5 CS6 · 3.1–3.5 CS6 · 3.6 CS7 · 4.1 CS4/CS8 · 4.2 CS2/CS4 · 4.3 CS4 · 4.4 CS4 · 4.5 CS9 · 4.6 CS7/CS9 · 5.1 CS5 · 5.2 CS5/CS8 · 5.3 CS3 · 5.4 CS6 · 5.5 CS9 · 5.6 CS3.

AIP-C01 (20/20): 1.1 CS1 · 1.2 CS3/CS1/ML2 · 1.3 CS3/CS4/ML1 · 1.4 CS4 · 1.5 CS4 · 1.6 CS5/CS8/CS4 · 2.1 CS2/CS3 · 2.2 ML3/CS1 · 2.3 CS1/CS8/CS7 · 2.4 CS2/CS3/CS9 · 2.5 CS7/CS3/CS9 · 3.1 CS8/CS4 · 3.2 CS8 · 3.3 CS8 · 3.4 CS8 · 4.1 CS9 · 4.2 CS9/CS4 · 4.3 CS9 · 5.1 CS9/CS4 · 5.2 CS5/CS4/CS9.

MLA-C01 (12/12): 1.1 ML1 · 1.2 ML1 · 1.3 ML1/CS8 · 2.1 ML1 · 2.2 ML2 · 2.3 ML2 · 3.1 ML3/CS9 · 3.2 CS1/CS6 · 3.3 ML3/CS7 · 4.1 ML3 · 4.2 CS9 · 4.3 CS8.

DEA-C01 (subset): 1.1 DE1 · 1.3 DE2 · 2.1 DE1 · 2.2 DE1 · 2.3 DE1 · 2.4 DE1 · 3.1 DE2 · 3.2 DE2 · 3.4 DE1 · (absorbed: 1.2→ML1; 1.4→CS6/CS7; 3.3→CS9; 4.1–4.5→CS8).


  1. Three-layer case template (Case / Assignment / Teaching Note).
  2. CS1 reference case (opener; reuses business-architecture diagrams as exhibits).
  3. DE1 second reference (bottom of stack; closest to real-world practice).
  4. Outward from there; CS10 capstone last.