A comprehensive AI-powered toolkit for the Mifos community, featuring intelligent chatbots, automation tools, translation utilities, and voice-driven banking interfaces. This project aims to provide self-service support, enhance developer productivity, and simplify access to Mifos ecosystem resources through advanced Generative AI technologies.
- Overview
- Key Components
- Quick Start
- Project Structure
- Technologies Used
- Contributing
- Links and Resources
As the Mifos community around its core products (Mifos X, Payment Hub EE, Mobile Applications, and Mifos Gazelle) grows rapidly, this repository provides a suite of AI-powered tools to support implementors, developers, and new users. The tools simplify finding relevant content across multiple sources, automate repetitive tasks, and enable innovative interfaces for banking applications.
Retrieval-Augmented Generation (RAG) chatbots for querying Mifos codebases using natural language.
| Bot | Notebook | Vector Storage | Description |
|---|---|---|---|
| Web App Bot | Web-App/web-app_bot.ipynb |
web_app_vector_storage_metadata/ |
Query the Mifos Web Application codebase |
| Mifos Mobile Bot | Mifos-Mobile/mifos-mobile_bot.ipynb |
mifos-mobile_vector_storage/ |
Explore Mifos Mobile app code and features |
| Android Client Bot | Android-Client/android-client_bot.ipynb |
android-client_vector_storage/ |
Navigate the Android Client codebase |
| Mobile Wallet Bot | Mobile-Wallet/mobile-wallet_bot.ipynb |
mobile_wallet_vector_storage/ |
Query Mobile Wallet implementation |
Note: Mifos Mobile also includes an alternative implementation (
mifos-mobile_bot_hf_groq.ipynb) using Groq LLM and Hugging Face embeddings instead of OpenAI.
Access hosted versions on Hugging Face Spaces:
- Processes and indexes Mifos codebases, creating vector embeddings using ChromaDB
- Uses OpenAI embeddings (text-embedding-3-large) for semantic search
- Leverages GPT models to generate human-like responses based on retrieved code snippets
- Provides insights on file organization, key components, and project architecture
# Clone the repository
git clone https://github.com/openMF/community-ai.git
cd community-ai
# Install dependencies
pip install -r requirements.txt
# Set up API keys in .env file
echo "OPENAI_API_KEY=your_openai_api_key" > .env
# OR
echo "GEMINI_API_KEY=your_gemini_api_key" >> .env
# Launch Jupyter Notebook
jupyter notebook
# Open any bot notebook (e.g., Web-App/web-app_bot.ipynb)
# Run cells and interact via Gradio interfaceFull-stack chatbot platform with FastAPI backend and Next.js frontend for integrating with Jira, Slack, and GitHub.
📂 Location: community-chatbot/
- Jira Agent: Query Jira issues using natural language, generate JQL, and summarize results
- Slack Agent: Manage conversations and summarize Slack channel discussions
- GitHub Agent: Interact with repositories and ask questions about project code
- Modern UI: React-based chat interface with authentication (Firebase)
- Agent-Based Workflows: LangChain-powered intelligent automation
- Backend: FastAPI (Python), LangChain, OpenAI
- Frontend: Next.js 14, React, Tailwind CSS, shadcn/ui
- Authentication: Firebase Auth
- Database: Firestore
cd community-chatbot
# Backend setup
python3 -m venv venv
source venv/bin/activate # On Windows: .\venv\Scripts\activate
pip install -r requirements.txt
# Configure environment variables
# Create a .env file with your API keys
# See how_to_get_keys.md for detailed setup instructions
# Add your API keys for OpenAI, Slack, Jira, GitHub, Firebase
# Start backend (from community-chatbot root)
bash scripts/run_backend.sh
# Or run individual agents:
# python scripts/jira.py
# python scripts/slack.py
# python scripts/github_agent.py
# Frontend setup (new terminal, from community-chatbot root)
pnpm install
pnpm dev
# Access at http://localhost:3000📂 Location: tools/translation-helper/
Gradio-based translation tool powered by Gemini 2.0 Flash.
Features:
- Supports 35+ languages (Arabic, Bengali, Chinese, French, German, Hindi, Japanese, Spanish, etc.)
- Formal/informal tone switching
- Real-time translation with Enter key or button trigger
cd tools/translation-helper
pip install -r requirements.txt
python app.py📂 Location: Mifos-Mobile Language Translation/
Translates Android strings.xml files using Groq LLM API.
Features:
- Parses
values/strings.xmland generates localized versions - Handles
<string>and<string-array>elements - Translation caching to avoid redundant API calls
- Safe XML parsing with pretty-printing
cd "Mifos-Mobile Language Translation"
python script.py📂 Location: WebApp Language Translations/
CLI utility to translate JSON string values using Groq LLM.
Features:
- Walks JSON structure and extracts string leaves
- Persistent caching per target language
- Exponential backoff for API reliability
cd "WebApp Language Translations"
python Script.py📂 Location: Voice-Driven_banking-Lam/, voice_driven_banking/
End-to-end voice banking assistant with React frontend and FastAPI backend.
- Multilingual Support: English, Hindi, Spanish, French, German, and more
- Core Banking Intents:
- Check account balance
- List recent transactions
- Transfer money with OTP verification
- Stateful Conversations: Context-aware multi-turn dialogues
- Animated UI: Real-time feedback for recording, processing, and speaking states
- AI Pipeline:
- STT: Whisper (Hugging Face Transformers)
- NLU/NLG: Google Gemini API (Vertex AI)
- TTS: MMS (Hugging Face Transformers)
- Backend: FastAPI, Google Gemini, Firestore
- Frontend: React, Tailwind CSS
- Audio Processing: FFmpeg, Whisper, MMS
# Backend setup
cd Voice-Driven_banking-Lam/Backend
python -m venv .venv
source .venv/bin/activate # On Windows: .\.venv\Scripts\Activate.ps1
pip install -r requirements.txt
# Configure .env file (see Backend README for details)
uvicorn main:app --reload
# Frontend setup (new terminal)
cd Voice-Driven_banking-Lam/Frontend/voicedriven
npm install
npm start📂 Location: voice_driven_banking/
Automated testing framework for voice banking interfaces.
cd voice_driven_banking
pip install -r requirements.txt
python voice_banking_test_suite.py📂 Location: Data Scraping/
Scrapes and extracts data from Mifos Jira/Confluence pages.
cd "Data Scraping"
jupyter notebook DataExtraction.ipynb📂 Location: Repo Clone Automation/
Automates cloning and processing of GitHub repositories for RAG applications.
Features:
- Selenium-based automated repo download
- Converts repository content to text for vector database ingestion
cd "Repo Clone Automation"
python repo_cloner.py
jupyter notebook github_repo_rag.ipynb📂 Location: Slack_scraper_bot/, slack_pipeline/
Tools for extracting, processing, and querying Slack messages.
- Message Extraction: Convert Slack JSON exports to clean text
- PII Removal: Automatically redact personal identifiable information
- RAG Integration: Query Slack conversations using vector databases
- Trustworthy LLM: Confidence scoring for generated responses
# Extract messages from Slack JSON
cd Slack_scraper_bot/scripts
python extract.py
# Remove PII
python pii_remocval.py
# Create vector database
cd ../../slack_pipeline
python main.pycd Slack_scraper_bot/scripts
jupyter notebook demo_bot.ipynb- Python 3.8+
- Node.js 18+ (for Next.js projects)
- Jupyter Notebook
- API Keys: OpenAI, Gemini, Groq (depending on components used)
# Clone repository
git clone https://github.com/openMF/community-ai.git
cd community-ai
# Install Python dependencies
pip install -r requirements.txt
# Set up environment variables
# Create a .env file and add your API keys
# OPENAI_API_KEY=your_key_here
# GEMINI_API_KEY=your_key_here
# GROQ_API_KEY=your_key_here
# Launch Jupyter for chatbots
jupyter notebook
# Or run specific tools (see component sections above)community-ai/
├── Android-Client/ # Android Client RAG chatbot
│ ├── android-client_bot.ipynb
│ └── android-client_vector_storage/
├── community_chatbot/ # Agent scripts (GitHub, Jira, Slack)
│ └── agent/
├── community-chatbot/ # Full-stack chatbot platform
│ ├── app/ # Next.js frontend pages
│ ├── components/ # React UI components
│ ├── scripts/ # FastAPI backend agents
│ └── package.json
├── Data Scraping/ # Jira/Confluence data extraction
│ └── DataExtraction.ipynb
├── Mifos-Mobile/ # Mifos Mobile app chatbot
│ ├── mifos-mobile_bot.ipynb
│ ├── mifos-mobile_bot_hf_groq.ipynb # Alternative using Groq/HF
│ └── mifos-mobile_vector_storage/
├── Mifos-Mobile Language Translation/ # Android XML translator
│ ├── script.py
│ └── Readme.md
├── Mobile-Wallet/ # Mobile Wallet chatbot
│ ├── mobile-wallet_bot.ipynb
│ └── mobile_wallet_vector_storage/
├── Repo Clone Automation/ # GitHub repo automation
│ ├── repo_cloner.py
│ └── github_repo_rag.ipynb
├── Slack_scraper_bot/ # Slack message extraction & bot
│ ├── scripts/
│ └── data/
├── slack_pipeline/ # Slack vector database pipeline
│ ├── main.py
│ └── vectordb.py
├── tools/
│ └── translation-helper/ # Multilingual translation tool
│ └── app.py
├── Voice-Driven_banking-Lam/ # Voice banking assistant
│ ├── Backend/
│ └── Frontend/
├── voice_driven_banking/ # Voice banking test suite
│ └── voice_banking_test_suite.py
├── Web-App/ # Web App chatbot
│ ├── web-app_bot.ipynb
│ └── web_app_vector_storage_metadata/
├── WebApp Language Translations/ # JSON translation utility
│ └── Script.py
├── CodeCommentingScript.ipynb # Code preprocessing utility
├── requirements.txt # Python dependencies
└── README.md
- LLMs: OpenAI GPT-4, Google Gemini 2.0 Flash, Groq
- Embeddings: OpenAI text-embedding-3-large
- Frameworks: LangChain, Hugging Face Transformers
- Vector Databases: ChromaDB, Pinecone
- Python: FastAPI, Flask
- Data Processing: BeautifulSoup, Selenium, pandas
- React/Next.js 14: Server-side rendering, App Router
- UI Libraries: Tailwind CSS, shadcn/ui, Radix UI
- State Management: Zustand
- Forms: React Hook Form, Zod validation
- Firebase: Firestore, Firebase Auth
- Vector Stores: ChromaDB (persistent), Pinecone
- STT: Whisper (Hugging Face)
- TTS: MMS (Hugging Face)
- Processing: FFmpeg
- Notebooks: Jupyter
- UI Prototyping: Gradio
- Testing: Jest, Selenium
- Package Management: pip, pnpm
We welcome contributions from the community! Here's how you can help:
- 🐛 Report bugs by opening issues
- 💡 Suggest new features or improvements
- 📝 Improve documentation
- 🔧 Submit pull requests with bug fixes or new features
- 🧪 Test and provide feedback on existing tools
-
Fork the repository
git clone https://github.com/YOUR_USERNAME/community-ai.git cd community-ai -
Create a feature branch
git checkout -b feature/your-feature-name
-
Make your changes
- Follow existing code style and conventions
- Add tests if applicable
- Update documentation
-
Commit your changes
git add . git commit -m "feat: add your feature description"
-
Push and create a Pull Request
git push origin feature/your-feature-name
Then open a PR on GitHub with a clear description of your changes.
- Use meaningful commit messages (follow Conventional Commits)
- Ensure code is well-documented with comments
- Test your changes before submitting
- Keep PRs focused on a single feature/fix
- Website: mifos.org
- Documentation: docs.mifos.org
- Jira: mifosforge.jira.com
- Slack: Join our AI Working Group
This project is licensed under the MPL-2.0 License - see the LICENSE file for details.
For questions, issues, or feature requests:
- 📧 Open an issue on GitHub
- 💬 Join our Mifos Slack community
- 📖 Check the documentation
Built with ❤️ by the Mifos Community