AI Platform Guides for Human-Centered Design with Agile
🎯 Important: These Are Meta-Prompt Templates
The files in this directory are NOT conversation starters. They are meta-prompt templates designed to be pasted into your AI agent as custom instructions or system prompts before you start your HCD-Agile work.
How to Use
- Choose your platform (OpenAI, Claude, Gemini, Copilot, or Grok)
- Load the template into your AI agent's custom instructions or at the start of your session
- Begin your actual work - the AI will now provide HCD-Agile guidance
See ReadMe.md for detailed usage instructions.
Overview
Human-Centered Design focuses on understanding and addressing real user needs through empathy, iteration, and validation. When combined with Agile methodologies and AI assistance, teams can:
- Conduct more thorough user research
- Synthesize insights more effectively
- Generate diverse solution ideas
- Prototype rapidly
- Test and iterate efficiently
Platform-Specific Guides
Each AI platform has unique strengths that can enhance different aspects of the HCD-Agile process. Choose the guide that matches your team's tools:
OpenAI (ChatGPT/GPT-4)
Best for: Conversational AI, code interpreter, custom GPTs
Key Strengths:
- Excellent general-purpose reasoning
- Strong at structured prompts and iterative refinement
- Code Interpreter for data analysis
- Custom GPTs for specialized workflows
Ideal for teams using: ChatGPT, OpenAI API, GPT-4
Anthropic Claude
Best for: Extended context, detailed analysis, XML-structured prompts
Key Strengths:
- 200K+ token context window (perfect for large research datasets)
- Excellent at nuanced, thoughtful analysis
- Strong at following complex multi-step instructions
- Projects feature for persistent context
Ideal for teams using: Claude, Claude API, Claude for Business
Google Gemini
Best for: Multimodal AI, long context, Google Workspace integration
Key Strengths:
- 1M+ token context window
- Multimodal understanding (text, images, video, audio)
- Native integration with Google Workspace
- Fast Flash model for real-time collaboration
Ideal for teams using: Gemini, Google AI Studio, Google Workspace
GitHub Copilot
Best for: Code-centric workflows, IDE integration, development automation
Key Strengths:
- Embedded in development workflow (VS Code, etc.)
- Code generation and completion
- Understanding of codebase context
- Rapid prototyping and test generation
Ideal for teams using: GitHub, VS Code, Visual Studio, JetBrains IDEs
xAI Grok
Best for: Real-time information, direct communication, trend analysis
Key Strengths:
- Real-time search and X/Twitter integration
- Understanding current trends and user sentiment
- Direct, conversational style
- Creative problem-solving with humor
- Access to up-to-date information
Ideal for teams using: Grok, xAI, X platform
The HCD-Agile Process
All guides follow the same five-phase HCD process adapted for Agile sprints:
1. 🔍 Discovery
Goal: Understand users and their needs
Activities:
- User research planning
- Interviews and surveys
- Observational studies
- Data collection and analysis
Sprint integration: Often spans multiple sprints; can be done incrementally
2. 📋 Define
Goal: Synthesize insights and define the problem
Activities:
- Research synthesis
- Affinity mapping
- User needs identification
- Problem statement creation
- "How Might We" questions
Sprint integration: Typically fits within one sprint; outputs inform backlog prioritization
3. 💡 Ideate
Goal: Generate diverse solution ideas
Activities:
- Brainstorming sessions
- Idea generation
- Concept exploration
- Solution prioritization
Sprint integration: Can be done in workshops within a sprint; generates backlog items
4. 🛠️ Prototype
Goal: Create testable representations of solutions
Activities:
- Rapid prototyping
- MVP definition
- Low-to-high fidelity prototypes
- Technical feasibility assessment
Sprint integration: Builds on sprint deliverables; outputs are sprint demos
5. ✅ Test
Goal: Validate solutions with users
Activities:
- User testing
- Feedback collection
- Iteration planning
- Success measurement
Sprint integration: Testing occurs at end of sprint; feedback informs next sprint
General Best Practices
Regardless of which AI platform you use:
1. Always Ground AI in User Research
- AI should assist, not replace, talking to real users
- Validate AI suggestions with user feedback
- Use AI to analyze and synthesize, not to assume user needs
2. Provide Rich Context
- Share user research findings with your AI assistant
- Explain project constraints and goals
- Reference user personas and problem statements
3. Iterate on AI Prompts
- First responses may be generic; refine with follow-up questions
- Be specific about format, depth, and perspective you need
- Build on previous responses in conversation
4. Balance Speed and Depth
- Use AI for rapid exploration, but don't skip critical thinking
- Agile values speed, but HCD requires some depth
- Time-box AI-assisted activities to stay within sprint constraints
5. Combine AI Strengths
- Use conversational AI for research and strategy
- Use code-focused AI for implementation
- Different tools for different phases and activities
6. Maintain the Human Element
- AI enhances human creativity and empathy, doesn't replace it
- Critical HCD activities (user interviews, testing) need human facilitation
- Use judgment in evaluating AI suggestions
7. Document Everything
- Keep track of research insights
- Document design decisions and rationale
- Maintain a clear audit trail for stakeholders
Sprint Integration Tips
Sprint Planning
- Include HCD activities as stories in your sprint
- Estimate HCD work like any other story (story points or time)
- Balance research/design with development work
Daily Standups
- Report on HCD activities alongside development progress
- Share user insights that might affect sprint direction
- Identify when user feedback is needed
Sprint Reviews
- Show prototypes and gather stakeholder feedback
- Present user research findings
- Demonstrate how work addresses user needs
Retrospectives
- Reflect on how well you understood and served users
- Evaluate effectiveness of HCD practices
- Improve the integration of HCD and Agile
Getting Started
- Choose your platform: Select the guide that matches your team's AI tools
- Identify your HCD phase: Determine where you are in the process
- Review relevant prompts: Find templates for your current activities
- Adapt to your context: Customize prompts with your project details
- Iterate and refine: Improve prompts based on results
Contributing
We welcome contributions to improve these guides! Consider adding:
- Platform-specific tips and tricks
- Real-world examples and case studies
- Prompt templates that worked well
- Integration patterns with other tools
- Common pitfalls and how to avoid them
Additional Resources
Human-Centered Design
Agile Methodologies
AI Prompt Engineering
- Platform-specific resources in each guide
- Prompt Engineering Guide
- Learn Prompting
Remember: AI is a powerful tool for HCD-Agile work, but it's most effective when used to enhance human creativity, empathy, and judgment—not replace them. Always validate AI-assisted work with real users.