Finsight RAG

It is a Streamlit application that pulls SEC filings (10-Q, 10-K) straight from EDGAR with no api, performs Mistral OCR on HTML/PDF documents, lets you chat with a Mistral powered LLM with streaming and RAG, and generates exportable reports.

1. Prerequisites

• Valid Mistral API key

Create and activate a virtual environment

python -m venv .venv
source .venv/bin/activate         # For Windows: .venv\Scripts\activate

Install dependencies

pip install -r requirements.txt

Install Playwright:

playwright install

2. Environment configuration

Place a .env file beside app.py or set variables in your UI:

MISTRAL_API_KEY=your-mistral-key

3. Run the application

streamlit run app.py

4. Feature overview

• Fetch & Index – Quickly grab filings by ticker, form, year, or quarter, or just upload a PDF/HTML yourself. The app breaks them into chunks and builds a smart search index.
• Chat – Ask questions. Mistral model streams back answers with extra reranking to make sure you get the most relevant info.
• Generate Report – Turn filings to markdown reports. You can export it straight to PDF.
• OCR Preview – Clean up messy HTML/PDF filings with Mistral OCR and get an easy to read markdown view with extracted images.

5. Compliance reminders

• Respect SEC guidance (~10 requests/second) and include your identity in every request.

6. Disclaimer

This application does not provide investment advice. The content is for informational and educational purposes only, aiming to assist users in exploring and understanding SEC filings.

7. Future Work

• Better reports
• Fix issue where exported images were not showing in Streamlit