Automated scraping and analysis pipeline for Malawi Stock Exchange financial documents.
# MSE AI Agent
**Last updated**: December 04, 2025
**Author**: Paul Namalomba
- SESKA Computational Engineer
- Software Developer
- PhD Candidate (Civil Engineering Spec. Computational and Applied Mechanics)
**Version**: v0.1.1 (Dec 2025)
**Contact**: kabwenzenamalomba@gmail.com
Automated scraping and analysis pipeline for Malawi Stock Exchange financial documents from `mse.co.mw`. It is an AI-powered agent that extracts financial data from PDFs, analyzes CEO statements, and provides structured data outputs for investor insights.
I have designed it to monitor, collect, and interpret financial data from the **Malawi Stock Exchange (MSE)**. The agent will automate the process of **scraping financial PDFs**, **storing structured data**, and **analyzing both quantitative and qualitative indicators** to estimate a company's fair stock price and future performance trends.
## Structure
- src/
- scraper/
- processing_ai/
- excel_writer/
- sql_writer/
- utils/
- data/
- financials/{COMPANY}/
## Quick start
1. Copy .env.example to .env and fill values.
2. Install requirements.
3. Run scraper.
## Notes
- **Scraper:**
- requests+BeautifulSoup4 for HTML scraping.
- OCR + Tesseract for PDF text extraction.
- respects robots.txt and rate limiting.
- **AI Agent:**
- Google Generative AI: Gemini 3 Pro.
- Extracts tables, CEO statements, KPIs.
- **Storage:**
- Initial version uses excel (.csv) exports.
- Pandas for Excel exports and data manipulation.
- Later: PostgreSQL for structured data.
---
**Current Version**: v0.1.0 (Nov 18 2025)
*Check CHANGELOG.md for updates.*