AI-DLC
AI-DLC: the AI-Driven Development Lifecycle
AI-DLC, the AI-Driven Development Lifecycle, is AWS's methodology for building software with AI agents doing the work and humans approving every important decision.
AI-DLC
The three phases
Inception
Requirements, planning and architecture. The AI asks clarifying questions, then proposes what to build and how, for a human to approve.
Construction
Design, code generation, build and test, one unit of work at a time, with a human approving each plan before code is written.
Operations
Deployment and monitoring, so what was built runs in production and stays healthy.
AI-DLC
How AI-DLC works
Instead of a human writing code with AI help, the AI drives each step: it asks clarifying questions, proposes a plan and waits for approval before it continues. Simple fixes skip straight to code; bigger features get requirements, design and testing stages. Every question, answer and decision is kept in an audit trail.
AI-DLC
AI-DLC with GitHub Copilot and Claude Code
AWS publishes AI-DLC as open-source workflows that run inside coding agents, including GitHub Copilot, Claude Code, Amazon Q Developer and Kiro. A team can adopt the lifecycle without changing tools.
AI-DLC pairs well with spec-driven development: the phases give the lifecycle and the approval points, and specs, plans and tasks give the day-to-day workflow inside each phase.
See how AI-DLC maps to agentic engineeringAI-DLC
FAQ
What is AI-DLC?
AI-DLC (AI-Driven Development Lifecycle) is AWS's methodology for AI-native software delivery. Work runs through three phases, Inception, Construction and Operations, where AI proposes plans and does the work while humans approve each step. It maps directly onto spec, plan, tasks and code.
Can I use AI-DLC with GitHub Copilot?
Yes. AWS's open-source AI-DLC workflows include setup for GitHub Copilot, so the lifecycle runs inside VS Code agent mode alongside your repository instructions and prompt files.
Is AI-DLC only for AWS?
No. It comes from AWS, but the method and its open-source workflows are not tied to AWS infrastructure and run in agents like GitHub Copilot and Claude Code.
Want help adopting AI-DLC?
I help engineering teams set up AI-DLC and spec-driven development on the tools they already use, from guidance files and gates to team training.