Summary

FinnewsHunter is an enterprise-grade, multi-agent financial intelligence system built on the AgenticX framework. It addresses the need for real-time financial news monitoring, deep interpretation, and sentiment analysis to generate alpha signals for quantitative trading. The system targets quantitative traders and financial analysts seeking automated insights from diverse news sources.

How It Works

The system employs a multi-agent architecture, integrating core AgenticX abstractions like Agents, Tools, and Workflows. A NewsAnalyst agent automatically processes news for sentiment and market impact using LLMs. Data flows from crawlers monitoring financial news sites, through LLM and embedding services, into storage layers (PostgreSQL, Milvus, Redis) for analysis and knowledge graph integration (planned). This approach enables deep interpretation beyond traditional text classification, aiming to uncover investment opportunities and risks.

Quick Start & Requirements

• Prerequisites: Python 3.11+, Docker & Docker Compose (optional), OpenAI API Key or local LLM.
• Installation: Clone AgenticX, install FinnewsHunter backend dependencies (pip install -r requirements.txt), configure .env with API keys, and run ./start.sh or docker-compose. Initialize the database (python -m app.core.database) and run the backend (uvicorn).
• Access: Backend API at http://localhost:8000, API Docs at http://localhost:8000/docs, Frontend at frontend/index.html.

Highlighted Details

• AgenticX Native: Deep integration with the AgenticX framework.
• Intelligent Agents: Features a NewsAnalyst agent for automated sentiment and market impact analysis.
• Comprehensive Tech Stack: Utilizes FastAPI, PostgreSQL, Milvus, Redis, and React.
• Production Readiness: Supports Docker Compose for one-click deployment, including logging and monitoring.
• Data Sources: Monitors multiple financial news outlets like Sina Finance, National Business Daily, and Securities Times.

Maintenance & Community

The project roadmap outlines phases for MVP completion, multi-agent collaboration (Debate workflow), knowledge enhancement (financial knowledge graph), and self-evolution (ACE framework integration). No specific community channels (e.g., Discord, Slack) are listed.

Licensing & Compatibility

The project states it follows the "AgenticX license." The specific terms and compatibility for commercial use are not detailed, requiring further investigation into the AgenticX license itself.

Limitations & Caveats

The current implementation represents Phase 1 (MVP), with core infrastructure, database models, and a basic NewsAnalyst agent complete. Advanced features such as multi-agent debate workflows, financial knowledge graph integration, and self-evolution capabilities are planned for future development phases. The frontend is described as a simplified version.