The capstone: build, validate, deploy, fail over, and gate a release.

Module 14 of 16 · track Ship · Estimated time: 8–10 hours · Prerequisites: Modules 11 and 13.

Everything converges here. You take TravelMind from source to a gated release — containerised, deployed, routed through a gateway, with model failover, a version manifest and a release pipeline that can say no.

Open the module on GitHub ↗: the slides, notebooks, exercises and solutions are there. This page is the reading order, with the tutorial and the workbench alongside.

The lesson that frames this module is Shadow Mode and Canary Releases for AI Agents, Step by Step (7 min): shadow, canary and cut-over, with the evidence each needs. Then try the cut-over evidence calculator in the workbench: the same decision, with your numbers.

• Package an agent for deployment with a reproducible build
• Route requests through a gateway with model failover
• Run a release pipeline with a readiness checklist and version manifest
• Diagnose and recover from a gate failure
• Present the whole architecture to a stakeholder

Portable GenAI concepts	AWS-specific surface
Release gating	AgentCore Runtime
Failover strategy	AgentCore Gateway
Versioning agents	IAM policies
Production readiness	CloudWatch
	Container builds
	Bedrock cross-region

Portable concepts transfer to any stack. The AWS column is where this module touches the cloud — see docs/concepts/portability-matrix.md.

#		Step	What it is
1	📖	slides/EndToEnd_Production.pptx	End-to-end production
2	📖	assets/reference_architecture.png	The reference architecture
3	💻	notebooks/TravelMind_Build_with_Decisions.ipynb	Build, with the decisions made explicit
4	📖	src/travelmind_agent.py	The agent source
5	📖	src/model_failover.py	Model failover
6	💻	notebooks/gateway_routing.ipynb	Gateway routing
7	💻	notebooks/TravelMind_Validate_with_Decisions.ipynb	Validate
8	💻	notebooks/deploy_e2e.ipynb	Deploy end to end
9	📊	src/readiness_checklist.md	Readiness checklist
10	📖	src/release_pipeline.md	Release pipeline
11	✏️	exercises/Exercise_6_Production_and_Gate_Failures.md	Production and gate failures
12	✏️	exercises/Capstone_RailReserve.md	Capstone — RailReserve

📖 read   💻 run   ✏️ practise   📊 workbook   🔖 reference

Worked solutions live in solutions/. Attempt every exercise closed-book first — the solutions are written to be read after you have a wrong answer to compare against.

• A deploy with no rollback path. Version the manifest, not just the image.
• Failover that silently degrades quality. Log which model actually answered.

Every error anyone has hit in this course is in the Error Index, by the exact string. Model access and region problems are in Model and Region Notes. If it is a question rather than an error, check the answered ones first, then ask in Q&A.

Frameworks

• Demo To Production Gap — Nine deltas, as a pre-mortem
• Reversibility Test — Rehearse the rollback, then time it
• Agent Readiness Scorecard — Seven dimensions before you ship

Quick reference

• Observability — What to instrument before launch

Recipes and procedures

• Runbook · rolling back — The four things that roll back together
• Runbook · first 30 days in production — A schedule, window by window

---

← Module 13 · Agentic QA and Evaluation · All modules · Module 15 · Agentic Product Lifecycle →

_{✏️ This page is generated from modules/14-end-to-end-production/README.md. To change it, edit the module README: an edit made here is replaced at the next sync.}