Skip to content

CyrilDesch/crewai-brainstormer

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🚀 CrewAI-Brainstormer - Mobile App Ideas Generator with Design Thinking of IBM

Design Thinking CrewAI Python

📋 Overview

CrewAI-Brainstormer is a powerful application that leverages the crewAI framework to automate the design thinking process, from IBM, and generate innovative mobile app ideas. The system simulates a collaborative team of AI agents, each with specialized roles, working through a structured design thinking methodology to create compelling product concepts with Hills statements.

CrewAI-Brainstormer autonomously creates innovative mobile app ideas by leveraging IBM's Enterprise Design Thinking toolkit with no user input required. The system generates random user personas, then simulates a complete design thinking workshop where AI agents collaborate through empathy mapping, scenario analysis, ideation, and solution refinement. The entire process - from random persona creation to final Hills statements - is handled by specialized AI agents working together to produce thoroughly researched mobile app concepts with clear value propositions.

Each generation process costs approximately $0.07 with Gemini 2.0 Flash.

🎯 Key Features

  • Multi-Agent Architecture: Utilizes specialized AI agents (UX Designer, Product Manager, Business Strategist, etc.)
  • AI-Powered Design Thinking: Automates the entire design thinking process from empathy mapping to solution design
  • Structured Workflow: Sequential process with validation, challenges, and refinement at each stage
  • True Randomization: Integrates with Random.org API for genuine randomization in persona creation
  • Parallel Execution: Run multiple idea generation instances simultaneously
  • Comprehensive Outputs: Generates complete reports with user personas, empathy maps, scenario maps, and Hills statements

🛠️ Requirements

  • Python >=3.10 and <3.13
  • OpenRouter API key - A unified API that routes to various AI models including OpenAI, Anthropic, and others
  • Random.org API key - Required for creating truly random user personas and preventing repetitive outputs

🚀 Getting Started

Setup

  1. Clone the repository:

    git clone git@github.com:Cyril-Deschamps/crewai-brainstormer.git
    cd crewai-brainstormer
  2. Use the Makefile to initialize the project:

    make init

    This will install all required dependencies including crewAI and prepare the environment.

  3. Configure your environment:

    • Copy .env.example to .env
    • Add your OpenRouter and Random.org API keys to the .env file

Running the Application

To generate one idea:

make run

To run multiple generation processes in parallel:

make run i=3  # Runs 3 instances in parallel

Viewing Results

To view all generated solution ideas and their Hills statements:

make solutions

🧠 How It Works

The CrewAI-Brainstormer follows the IBM Design Thinking methodology to create innovative solutions:

  1. Initialization: Creates random personas with specific needs and problems
  2. Empathy Mapping: Develops deep understanding of user needs through simulated interviews
  3. As-Is Scenario Mapping: Documents current workflow and pain points
  4. To-Be Scenario Mapping: Envisions an ideal future state
  5. Big Idea Vignettes: Creates compelling narratives around the solution
  6. Hills Statements: Defines the "Who-What-Wow" of the solution

Multiple AI agents collaborate through these stages with built-in challenges and verification steps to ensure quality.

📂 Project Structure

crewai-brainstormer/
├── src/
│   ├── config/
│   │   ├── agents.yaml  # Agent roles and personalities
│   │   └── tasks.yaml   # Task definitions and workflows
│   ├── CrewAIBrainstormer.py  # Main crewAI implementation
│   ├── main.py         # Entry point
│   ├── random_helper.py # Random.org integration and utilities
├── scripts/            # Utility scripts for project management
├── conversations/      # Output directory for reports
├── .env.example       # Example environment variables
├── Makefile           # Project automation
├── pyproject.toml     # Project metadata and dependencies
└── uv.lock            # Dependency lock file

📝 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgements

About

A multi-agent system based on AI to follow the Design Thinking process of IBM.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors