A multi-agent AI system that simulates strategic debate to help you make better decisions. Upload your context (Knowledge Base), ask a question, and Optimist, Pessimist, and Synthesizer agents will communicate to provide a balanced recommendation.
Most AI tools give you either overly optimistic or generic advice. This system forces a structured discussion between multiple perspectives, ensuring you get a more realistic and nuanced analysis of your strategic questions.
- Frontend: React with TypeScript (Next JS + Tailwind)
- Backend: FastAPI (Python)
- AI Models: Ollama (Mistral) for local inference
- Embeddings: Sentence Transformers for semantic search
- Storage: File-based knowledge base (txt/json)
- Upload relevant documents to the knowledge base
- Ask strategic questions through the UI
- Three AI agents independently analyze the question:
- Optimist: Identifies opportunities and growth paths
- Pessimist: Highlights risks and challenges
- Synthesizer: Creates a balanced recommendation
- Receive a final recommendation that considers all viewpoints.
- Real-time loading of each perspective
- Context-aware responses using RAG
- Balanced synthesis of conflicting viewpoints
- Support for text and JSON knowledge files