Guide
What Is AIDLC (AI-DLC)? The AI-Driven Development Lifecycle
Last updated: October 27, 2026
AI-DLC, also written AIDLC, stands for AI-driven development lifecycle. It is a way to build software where AI does most of the planning and coding work, and people make the decisions. AWS introduced the name in 2025 and published its workflows as open source.
The idea is simple. In a classic SDLC, people do the work and tools help. In AI-DLC, AI proposes the work and people review, correct and approve it.
The three phases of AI-DLC
- Inception. AI turns a business intent into requirements, user stories and units of work. It asks clarifying questions, and the team answers and corrects them together.
- Construction. AI proposes the design, then the code and tests. The team reviews each step and overrides what is wrong.
- Operations. Deployment and monitoring. This is the least mature phase, and most teams keep it under direct human control for now.
AI-DLC vs. the traditional SDLC
The SDLC was designed around people writing code, so its cycles are measured in weeks. AI-DLC assumes the writing is cheap and the deciding is not. Cycles shrink to hours or days, and the scarce resource becomes the reviewer's attention.
That changes what matters most. A clear spec matters more than a fast typist. A visible plan matters more than a status meeting. Every AI decision needs a record, so a person can check why it was made.
What makes AI-DLC work
Approval before action. The AI writes a plan and waits. Nothing is built until a person agrees with the plan.
Questions, not guesses. When the intent is unclear, the AI asks. Answers are written down and become part of the spec.
An audit trail. Each step, from requirement to pull request, can be traced back to the decision that caused it.
Running AI-DLC with nTra
nTra is a spec-driven product engineering platform: every task your AI coding agents build traces back to a requirement your business signed off. It is built on the same principles. Inception happens in Discovery: nTra interviews you, writes the brief, and flags the gaps. You approve the brief, and nTra breaks it into documents, screens and sized tasks.
Construction happens on the Board and in the Console. Each task runs through your own coding agent, such as Claude Code, Codex or Copilot. nTra re-runs the checks itself and ends each task as a pull request. Merging and production releases stay with your team.
See the full loop on how nTra works, or read about spec-driven development. For the general term, see AI SDLC.