This project implements an autonomous multi-agent system designed for real-time competitive analysis and strategic insight generation. By orchestrating a suite of specialized AI agents, the platform transforms raw market noise into actionable intelligence, providing users with a strategic advantage. It leverages an event-driven architecture and a microservices approach to efficiently gather, process, analyze, and report on competitive landscapes.
- Autonomous Orchestration: A central Orchestrator agent manages and coordinates the workflow of specialized AI agents through an event-driven (Pub/Sub) system.
- Data Acquisition (Fetcher Agent): Gathers unstructured market data from diverse sources (web, APIs) to build a comprehensive view of competitors.
- Data Vectorization (Embedding Service): Transforms raw data into high-dimensional vector embeddings, enabling advanced semantic search and analysis.
- Semantic Analysis (Analyzer Agent): Utilizes large language models (LLMs) like Gemini 2.0 (via
gemini-2.0-flash) and vector embeddings to extract deep, actionable insights from processed data. - Quality Assurance (Evaluator Agent): Ensures the accuracy and relevance of agent outputs against defined criteria. (Automatically invoked by the Analyzer).
- Automated Reporting (Reporter Agent): Compiles validated intelligence into structured, human-readable reports (PDF, JSON) for stakeholders.
- Real-time Notifications (Notifier Agent): Delivers critical alerts and report completion notifications via various channels.
- Scalable Microservices Architecture: Built on Google Cloud Platform (GCP) with FastAPI for backend services and Next.js for a responsive frontend.
- Secure Authentication: Integrates with Clerk.dev for robust user authentication and authorization.
The project is built with a modern, scalable technology stack, designed for performance and ease of development.
- Python 3.x: The primary programming language for all backend services and agents.
- FastAPI: A modern, fast (high-performance) web framework for building APIs with Python 3.7+ based on standard Python type hints.
- SQLAlchemy: SQL toolkit and Object-Relational Mapper (ORM) that provides a flexible and powerful interface to relational databases.
- PostgreSQL: A powerful, open-source object-relational database system.
- Google Cloud Platform (GCP):
- Pub/Sub: Used for asynchronous communication and event-driven architecture between agents.
- Cloud Trace: For distributed tracing and observability of microservices.
- Cloud Storage: For storing raw data, processed insights, and final reports (JSON, PDF).
- Vertex AI: Google Cloud's machine learning platform, utilized for:
- Large Language Models (LLMs): Gemini 2.0 (via
gemini-2.0-flash) for advanced natural language understanding and generation. - Embeddings: For converting text into numerical vector representations.
- Vector Search: For efficient similarity search on embedded data.
- Large Language Models (LLMs): Gemini 2.0 (via
- Pydantic: Data validation and settings management using Python type hints.
- Tenacity: General-purpose retrying library.
- Svix: Webhook delivery as a service.
- Next.js: A React framework for building performant, server-rendered and static web applications.
- TypeScript: A typed superset of JavaScript that compiles to plain JavaScript.
- Tailwind CSS: A utility-first CSS framework for rapidly building custom designs.
- Clerk.dev: For secure and seamless user authentication and management.
- react-hot-toast: A lightweight and customizable toast notification library.
- React Flow: A library for building node-based editors and interactive diagrams, used for visualizing orchestration graphs.