Your AI Co-Pilot for Product Management

ProdigyPM is an Agentic AI platform that helps Product Managers plan, ideate, research, and automate workflows using multi-agent AI orchestration powered by NVIDIA Nemotron and local LLMs.

• 7 Specialized Agents working in harmony:
  • StrategyAgent: Market sizing, idea generation, competitive analysis
  • ResearchAgent: User research, trend analysis, sentiment analysis
  • DevAgent: User stories, backlog generation, technical specs
  • PrototypeAgent: Design mockups, Figma integration
  • GtmAgent: Go-to-market strategy, launch planning, pricing
  • AutomationAgent: Sprint summaries, workflow automation
  • RegulationAgent: Compliance checks, risk assessment (PNC challenge)

• Multi-Agent Orchestration with LangGraph
• NVIDIA Nemotron Integration for strategic reasoning (budget under $40)
• Local LLM Support via Ollama (Llama 3 8B or Mistral)
• Vector Memory with FAISS for context-aware agent communication
• Real-time WebSocket Updates for live agent status
• Integration Stubs for Jira, Slack, Figma, Reddit

• Futuristic Design with clean, modern aesthetic
• Framer Motion Animations for smooth transitions
• TailwindCSS with custom color palette
• Real-time Dashboard with agent panels and activity feed
• Chat Interface for natural language interaction
• Analytics & Insights page with charts and metrics

ProdigyPM/
├── backend/                   # FastAPI Backend
│   ├── agents/               # 7 AI Agents
│   │   ├── base_agent.py
│   │   ├── strategy_agent.py
│   │   ├── research_agent.py
│   │   ├── dev_agent.py
│   │   ├── prototype_agent.py
│   │   ├── gtm_agent.py
│   │   ├── automation_agent.py
│   │   └── regulation_agent.py
│   ├── orchestrator/         # Multi-agent coordination
│   │   ├── task_graph.py    # LangGraph workflow
│   │   ├── memory_manager.py # FAISS vector store
│   │   └── nemotron_bridge.py # NVIDIA Nemotron API
│   ├── integrations/         # External APIs (mock for MVP)
│   │   ├── jira_api.py
│   │   ├── slack_api.py
│   │   ├── figma_api.py
│   │   └── reddit_api.py
│   ├── db/                   # SQLite database
│   │   └── context_store.py
│   ├── utils/                # Configuration & logging
│   └── main.py              # FastAPI app with WebSocket
│
└── frontend/                 # React + TypeScript Frontend
    ├── src/
    │   ├── components/      # React components
    │   │   ├── Dashboard.tsx
    │   │   ├── AgentPanel.tsx
    │   │   ├── TaskCard.tsx
    │   │   ├── ChatInterface.tsx
    │   │   └── ReportView.tsx
    │   ├── pages/           # Page components
    │   │   ├── Home.tsx
    │   │   ├── ProjectDashboard.tsx
    │   │   └── Insights.tsx
    │   ├── hooks/           # Custom React hooks
    │   │   └── useAgents.ts
    │   ├── utils/           # API client
    │   │   └── apiClient.ts
    │   └── styles/          # TailwindCSS
    └── package.json

• Python 3.10+
• Node.js 18+
• Ollama (optional, for local LLM)
• NVIDIA API Key (optional, for Nemotron)

git clone https://github.com/yourusername/ProdigyPM.git
cd ProdigyPM

cd backend

# Create virtual environment
python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

# Create .env file
cp .env.example .env
# Edit .env and add your API keys (optional)

# Run the backend
python main.py

Backend will start on http://localhost:8000

cd frontend

# Install dependencies
npm install

# Create .env file
cp .env.example .env

# Run development server
npm run dev

Frontend will start on http://localhost:5173

Navigate to http://localhost:5173 to see ProdigyPM in action!

# Ollama Settings (for local LLM)
OLLAMA_BASE_URL=http://localhost:11434
OLLAMA_MODEL=llama3:8b

# NVIDIA Nemotron (for strategic reasoning)
NEMOTRON_API_KEY=your_nvidia_api_key_here
NEMOTRON_MAX_CALLS=3  # Budget control

# Integration APIs (optional - uses mock data if not provided)
JIRA_API_TOKEN=your_token
SLACK_BOT_TOKEN=your_token
FIGMA_ACCESS_TOKEN=your_token
REDDIT_CLIENT_ID=your_id
REDDIT_CLIENT_SECRET=your_secret

For local LLM processing:

# Install Ollama
curl https://ollama.ai/install.sh | sh

# Pull a model
ollama pull llama3:8b
# or
ollama pull mistral

1. Select a Workflow from the dropdown:

   • Full Feature Planning (all agents)
   • Research & Strategy
   • Dev Planning
   • Launch Planning
   • Compliance Check

2. Click "Run Workflow" to execute

3. Watch Agents Work in real-time via WebSocket updates

Type natural language queries:

• "Research user pain points for PM tools"
• "Generate user stories for AI dashboard feature"
• "Create a go-to-market plan for B2B SaaS"
• "Check compliance for financial data feature"

Navigate to the Insights page to see:

• Time saved metrics
• Agent performance analytics
• Task completion rates
• Activity charts

• Base: Charcoal #0F1117
• Accents:
  • Neon Cyan #00FFFF
  • Soft Orange #FF7A00
• Neutral Grays: #1A1D29 to #3A3E4A

• Headings: Orbitron
• Body: Inter
• Code: Fira Code

• RegulationAgent specializes in financial compliance
• Checks GDPR, SOC 2, PCI-DSS, SOX, GLBA
• Risk assessment and audit reports
• Privacy review capabilities

• Nemotron Integration for multi-step reasoning
• Strategic orchestration of multiple agents
• Budget-controlled API calls (max 3 per session)
• Fallback to local LLM when appropriate
• Caching and smart routing

# Deploy to Railway
railway init
railway up

# Or deploy to Render
# Connect your GitHub repo to Render
# Set environment variables in Render dashboard

# Deploy to Vercel
vercel deploy

# Or deploy to Netlify
netlify deploy

Make sure to set all required environment variables in your deployment platform.

Strategy:

• Nemotron: Limited to 2-3 calls per major workflow
• Ollama: Free local inference for routine tasks
• Caching: Response caching to avoid duplicate API calls
• Smart Routing: Use Nemotron only for strategic tasks

Estimated Costs:

• Nemotron API: $20-30 (50-100 calls)
• Hosting: Free tier (Railway/Render + Vercel)
• Total: Well under $40

Once the backend is running, visit:

• Swagger UI: http://localhost:8000/docs
• ReDoc: http://localhost:8000/redoc

• POST /api/v1/run_task - Execute multi-agent workflow
• POST /api/v1/agents/{agent_name}/execute - Run single agent
• GET /api/v1/agents - Get agent status
• WS /ws/agents - WebSocket for real-time updates
• GET /health - Health check

Contributions are welcome! Please follow these steps:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

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

• NVIDIA for Nemotron API access
• PNC for the productivity challenge inspiration
• Ollama for local LLM infrastructure
• LangChain/LangGraph for agent orchestration patterns
• FastAPI and React communities

• Email: contact@prodigypm.ai
• Twitter: @ProdigyPM
• Discord: Join our community

• Real Ollama LLM integration
• Production Nemotron API calls
• Real FAISS vector embeddings
• Actual integration APIs (Jira, Slack, etc.)
• User authentication
• Team collaboration features
• Mobile app
• Chrome extension
• More specialized agents

Built with care by the ProdigyPM Team

#AIforPMs #ProductManagement #NVIDIA #PNC #Hackathon2025