Skip to content

AI Development Workflow

This section outlines workflows for integrating AI into the development lifecycle. Start with project setup, then choose the workflow that best fits your project needs and team structure.

Getting Started

Project Setup - Define foundational elements, tools, and conventions (start here for all workflows)

Choose Your Development Workflow

We provide three complementary approaches for different project needs:

1. Feature-Based Development

Best for: Enterprise projects, complex systems, and team-based development requiring thorough documentation and process adherence.

Get Started:

Complete Process:

  1. Functional Requirements - Define expected behavior, UI, testable features, and user experience using AI tools to draft and refine
  2. Technical Specification - Define technical implementation, data models, API endpoints, and backend architecture
  3. Development - Develop the solution iteratively, leveraging AI for implementation
  4. Testing - Unit, functional and end-to-end testing
  5. Refactoring - Refactor once feature complete
  6. Documentation - Update all relevant documentation to reflect changes

Use when:

  • Building complex features with multiple components
  • Working across multiple teams or stakeholders
  • Documentation and knowledge preservation are critical
  • Following formal development processes

2. Spec-Driven Development

Best for: Systematic requirements management with AI-optimized planning before implementation.

Get Started:

  • Overview - Quick introduction and decision framework
  • Process Guide - Three-phase methodology from idea to implementation
  • Examples - Real specs and case studies

Three-Phase Process:

  1. Requirements - Transform vague ideas into clear, testable requirements
  2. Design - Create comprehensive technical plans for implementation
  3. Tasks - Break down designs into actionable, sequential implementation steps

Use when:

  • Features involve multiple components or integrations
  • Working on high-stakes projects where failure is costly
  • Coordinating work across multiple developers or teams
  • Using AI tools that benefit from structured context
  • Requirements are complex or likely to change

3. Vibe Coding Plus

Best for: Rapid prototyping, small projects, and situations where speed and flexibility are prioritized.

Get Started:

Use when:

  • Building experimental prototypes
  • Working on small, well-understood features
  • Time-critical development
  • Solo development or small teams
  • Iterating quickly on ideas

Additional Resources

Learning with AI - Continuous learning and skill development with AI assistance

Choosing the Right Workflow

Factor Feature-Based Spec-Driven Vibe Coding Plus
Project Size Large, complex Medium to large Small to medium
Team Size Multiple agile teams 1-10 developers single developers
Documentation Comprehensive Structured Minimal
Planning Time Extensive Moderate Minimal
Risk Tolerance Low Low-Medium Medium-High
Change Frequency Low Medium High

Not sure which to choose? Start with Project Setup, then try Spec-Driven Development for your first AI-assisted feature to learn the principles before adapting to your team's needs.