AIUP (AI Unified Process)

Overview

AIUP is an agile, iterative software development methodology that combines principles from the Rational Unified Process (RUP) with modern AI tooling. It keeps requirements at the center while using AI to generate specifications, code, and tests from those requirements. Rather than treating AI as an autonomous code generator, AIUP positions AI as a consistency engine that collaborates with human developers throughout the iterative development cycle.

Core Philosophy

Perfect specifications are neither possible nor necessary upfront. AIUP rejects the “determinism myth”—the belief that AI code generation only works with exhaustive specifications that force deterministic output. Instead, specifications, code, and tests improve together through continuous short iterations, with each cycle building on the previous one.

Six Foundational Principles

1. Requirements-driven - Development begins and remains anchored to explicit requirements
2. AI-assisted - AI tools help with generation across specifications, code, and tests
3. Iterative approach - Short cycles where specifications, code, and tests evolve together
4. Test-driven consistency - Tests ensure the system behaves consistently regardless of how AI generates code
5. Stakeholder-centric - Business people can directly revise system use cases and requirements documents
6. Collaborative disciplines - All disciplines work together across iterations rather than in sequential phases

Development Workflow

AIUP follows a structured, incremental sequence (not upfront):

1. Business Requirements Catalog - Manually developed by business stakeholders
2. Business Use Case Diagrams - AI-generated, then reviewed with stakeholders
3. Entity Models - AI-derived from requirements catalog and reviewed
4. System Use Case Diagrams - AI-generated and revised by developers/stakeholders
5. System Use Case Specifications - AI-generated detailed markdown, revised by team
6. Application Code - AI-generated and reviewed by developers

Each step flows through iterations rather than completing all requirements before development begins.

Operational Model

• Continuous short iterations where all disciplines work simultaneously rather than sequentially
• Artifact review: Stakeholders review every artifact to ensure system matches actual needs
• Living documentation: Documentation enables refactoring and modernization without losing organizational knowledge
• Consistent behavior assurance: Tests protect system behavior while AI improves code generation quality
• Complete traceability: Every connection from business requirement to code line is maintained

Role of Testing in AI-Driven Development

Tests serve as the critical safeguard in AIUP. Rather than relying on perfectly specified requirements upfront, tests ensure that:

• Regardless of how AI generates or refactors code, behavior remains consistent
• System reliability is maintained during evolution
• AI can safely improve code quality without risking functionality

Key Advantages

• Faster development: AI handles tedious work—diagrams, specifications, code generation
• Business focus: Stakeholders focus on requirements rather than implementation details
• Complete traceability: Full connection from business requirement to code line
• Maintainable code: Living documentation enables sustainable development and modernization
• Better alignment: Requirements remain source of truth, not code

AIUP vs. Other AI Development Approaches

AIUP represents a balanced approach:

• Not purely autonomous: AI doesn’t generate entire applications independently
• Not purely assisted: AI does more than enhance specific tasks like code completion
• Human-centric with AI execution: AI creates plans and seeks clarification; humans retain decision-making authority for business-critical choices

AIUP in the AI Development Landscape

AIUP exists between:

• AI-assisted development: AI enhances specific tasks (documentation, code completion)
• AI-autonomous development: AI generates entire applications
• AIUP (balanced): AI-powered execution with human oversight throughout

Integration with Claude Code

AIUP is implemented as a methodology within Claude Code through plugins that bring requirements-driven development into AI-powered environments. This enables seamless integration of AI generation with human review and validation.

Technology-Specific Implementation

AIUP supports technology-specific plugins:

• Java/Vaadin plugin: Combines Vaadin UI framework with jOOQ type-safe database access
• Community contributions welcome for additional technology stacks

Related Concepts

• Rational Unified Process (RUP) - Original process inspiration
• Claude Code - Primary implementation environment
• Domain-driven design
• Use case modeling
• Requirements management
• Agile methodology

---

Last updated: 2026-01-23