Finanalyser: Financial Analysis Multi-Agent Research Tool

Overview

This project is a multi-agent hierarchical financial analysis research tool built using LangGraph. It demonstrates how hierarchical multi-agent frameworks can be implemented for financial data analysis, research, and prediction.

The system consists of two primary nodes:

1. Data Fetch & Store Node
2. Q&A Node

Each of these nodes contains specialized sub-nodes to handle different types of financial data.

---

System Architecture

1. Data Fetch & Store Node

This node is responsible for fetching financial data from different sources and storing them efficiently.

It consists of two sub-nodes:

a. YFinance Fetching Node

• Fetches financial data such as stock prices, earnings reports, and historical trends using the YFinance API.
• Stores the structured financial data in a PostgreSQL database.

b. Financial Document Ingestion Node

• Accepts financial documents like balance sheets, income statements, and cash flow statements in PDF format.
• Extracts relevant data and stores it in a vector database (FAISS) for efficient retrieval.

---

2. Q&A Node

This node processes user queries by searching financial data across different sources.

It consists of three sub-nodes:

a. Vector Search Node

• Searches the vector database for relevant information from ingested financial PDFs.
• Uses FAISS to perform fast and efficient retrieval of stored document embeddings.

b. PostgreSQL Query Node

• Fetches financial data for a given company from the PostgreSQL database.
• Calls the Pandas Agent to analyze the retrieved data and generate responses.

c. Web Scraper Node

• Performs a real-time web search for additional financial data.
• Uses scraping APIs to extract relevant financial insights from the internet.

Each time a user asks a financial question, all three sub-nodes (Vector Search, PostgreSQL Query, Web Scraper) are executed, and the system returns results from each source.

---

Technology Stack

• Python (Main backend development)
• LangGraph (Hierarchical multi-agent framework)
• YFinance API (Stock market data retrieval)
• FAISS (Vector store for efficient document retrieval)
• PostgreSQL (Relational database for structured financial data)
• Web Scraping APIs (For real-time financial data extraction)

---

How It Works

1. User requests financial data → The system fetches data from YFinance and PostgreSQL.
2. User uploads financial documents → The system processes and stores them in a vector database.
3. User asks a financial question → The system retrieves relevant answers from:
   • Vector DB (PDFs)
   • PostgreSQL DB (Structured data + Pandas analysis)
   • Web Search (Real-time financial insights)

All three answers are returned, providing a comprehensive financial analysis.

---

This setup ensures that financial insights are data-driven, sourced from multiple locations, and cross-verified for accuracy. 🚀

Installation & Setup

1. Clone the repository:

   git clone https://github.com/your-repo-name.git
   cd your-repo-name

2. Create a virtual environment and install the requirements:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   pip install -r requirements.txt

3. Set up environment variables using the .env.sample as a reference.
4. To start the API, use the following command:

   uvicorn app.api.api:app --host 0.0.0.0 --port 8001

5. Once the server is running, you can make a request using:

   curl -X POST "http://localhost:8001/query/" -H "Content-Type: application/json" -d '{"query": "Fetch the data for Apple"}'