# BMAD: The AI-Driven Delivery Framework

> What BMAD is, how its persona-based workflow guides you from ideation to agentic implementation, and where it fits in the SDLC.

- **Source:** https://ainativesoftware.engineering/roadmap/day-6/bmad-the-ai-driven-delivery-framework
- **Site:** AI-Native Software Engineering — https://ainativesoftware.engineering/book

- **Day:** 6 · position 1 of 3
- **Reading time:** 2 minutes
- **Day overview:** [Day 6](https://ainativesoftware.engineering/roadmap/day-6.md)

What BMAD is, how its persona-based workflow guides you from ideation to agentic implementation, and where it fits in the SDLC.

## Theory

**BMAD** (Breakthrough Method of Agile AI-Driven Development) is a structured SDD framework built around specialized AI agent personas and guided workflows. It covers the full delivery lifecycle: from ideation and PRD creation through architecture, story breakdown, implementation, and verification.


### Core concepts

**Personas** BMAD defines role-based AI personas (PM, Architect, Developer, QA, Scrum Master) that each bring focused expertise to a specific phase of delivery.


**Guided workflows** Each workflow orchestrates a sequence of agent interactions that produce consistent, high-quality artifacts regardless of who runs them.


**Artifact-driven** BMAD produces a well-defined artifact set at each stage: project brief, PRD, architecture doc, stories with acceptance criteria, task lists. These artifacts are the source of truth for the AI agents that implement them.


### The BMAD delivery loop

1. **Ideation** Use the PM persona to refine a product idea into a structured PRD
2. **Architecture** Use the Architect persona to define technical architecture, API contracts, and component structure
3. **Story creation** Break work into stories with acceptance criteria and test scenarios
4. **Agentic implementation** AI agents implement each story against its spec
5. **Verification** Tests, linters, and review gates validate that output matches the spec


### Why teams adopt BMAD

- Consistency across engineers: everyone follows the same workflow, producing comparable artifacts
- Quality gates baked in: the framework doesn't allow skipping verification steps
- Scales to complex projects: BMAD handles brownfield codebases, multi-team coordination, and long-lived products where simpler approaches break down
- Cross-functional alignment: the PM and Architect personas bring non-engineering stakeholders into the spec-writing process

**Theory resources**

- [Getting Started with BMAD - Official Docs](https://docs.bmad-method.org/)
- [BMAD - Github](https://github.com/bmad-code-org/BMAD-METHOD)

## Practice

Run this in a repository you already know, not a toy project.

Read the BMAD docs, install it in a real project, and spend time exploring what the agents can do.

**Step 1: Read the docs**

Go to [docs.bmad-method.org](https://docs.bmad-method.org) and read the Getting Started section. Get a feel for the structure: what BMAD is, what it installs, and how it works inside your IDE.

**Step 2: Install BMAD**

Follow the installation guide to set up BMAD in a project you own (or create a new empty repo). Complete the installation and verify the rules files are in place.

**Step 3: Invoke the agents**

Open your AI coding tool and try loading a few BMAD agents (PM, Architect, Developer, Scrum Master). For each one, ask it to describe its role and what tasks it can help with.

**Step 4: Use the help command**

Type `/bmad-help` and explore the available workflows. Read through the list and note which ones would be most useful on the kind of projects you typically work on.

**Reflect:**
- Which agents felt immediately useful?
- Which workflows surprised you?
- What would you use BMAD for on your current project?

- **Previous topic:** [Spec-kit Under the Hood](https://ainativesoftware.engineering/roadmap/day-5/spec-kit-under-the-hood.md)
- **Next topic:** [The BMAD Workflow Map](https://ainativesoftware.engineering/roadmap/day-6/the-bmad-workflow-map.md)

---

_AI-Native Software Engineering by Alfonso Graziano (O'Reilly Media, Early Release; print edition February 2027). Every page of ainativesoftware.engineering is also served as Markdown: append `.md` to any URL. Index: https://ainativesoftware.engineering/llms.txt — whole site in one file: https://ainativesoftware.engineering/llms-full.txt._