From Spec to Production Code — Kiro, Claude Code & Codex on AWS
The demo shipped. The AI writes code. Everyone is impressed.
Then someone asks if it can run in production — with audit logs, IAM controls, and no model lock-in — and the room goes quiet.
That gap between "works on my machine" and "deployed with governance" is where most AI coding projects die. This workshop is where you close it.
What This Is
A 90-minute, two-part AWS masterclass at the AWS HCMC Office — built for engineers who are actively building at Agentic AI Build Week.
In 90 minutes, you will:
Build a structured spec with Kiro and understand why every AI project that skips this step pays for it later
Learn how Kiro Hooks enforce engineering standards automatically — continuously, not just at commit time
Deploy Claude Code and Codex on Amazon Bedrock with IAM, usage policies, and CloudTrail wired in from day one
Apply agentic coding patterns — multi-file refactoring, test generation, codebase reasoning — using natural language commands
Leave with a production-hardened agent spec: tested, monitored, and ready to ship
Part 1 — Spec-Driven Development with Kiro
Most AI projects fail before a line of code is written — because the spec never existed. Kiro enforces a discipline most teams skip: structured requirements → design contract → code. You'll walk the full lifecycle live, then learn how Hooks keep standards enforced at every step after.
Part 2 — AI-Native Coding with Claude Code & Codex on Amazon Bedrock
With the spec locked, you get 10x velocity without sacrificing control. Every model call stays inside your AWS boundary. Swap models without changing your workflow. Full logging and CloudTrail integration for auditability baked in — not bolted on. You'll apply agentic coding patterns to harden the agent spec from Part 1 using natural language commands.
Who This Is For
This workshop is for you if:
You're a developer, solutions architect, or technical founder building AI-assisted applications
You've used AI coding tools (Copilot, Cursor, or similar) and hit the ceiling on governance and control
You're building something at AABW that needs to be production-ready — not just demo-ready
This is NOT for you if you want another "prompt engineering 101" session. This is engineering discipline applied to agentic AI.
Your Speaker
Diep Nguyen is a Solutions Architect at Amazon Web Services, based in Vietnam and working across the ASEAN region. He specialises in helping development teams move from AI prototypes to production-grade systems on AWS — with governance, flexibility, and auditability built in from the start.
The Context
This session is part of Agentic AI Build Week (Jul 8–12) — Southeast Asia's largest agentic AI hackathon, hosted at the AWS HCMC Office with 2,000+ builders, 30+ tech partners, and a $1M+ prize pool.
Main event registration required. Apply here first: Register for AABW
Seats are limited to the AWS office capacity. Applications are reviewed — engineers actively building at AABW get priority.
👉 Apply below if you're shipping something real, not just watching.