Back to Portfolio
AI/ML2024

Risk Factors Visualization

Stanford Law School hackathon project automating risk factor analysis in SEC filings (10-K and 10-Q) using NLP and interactive graph visualization.

Role
Developer
Timeline
Apr 2024
Team
Greg Tanaka, Han-chung Lee, Giuseppina D'Auria, Yassin Kortam
Year
2024
Risk

Overview

Developed a Risk Factors Visualization Tool during Stanford Law School hackathon to automate risk factor analysis in SEC filings. Automated extraction and categorization of risk factors from legal documents with interactive graph visualization.

The Challenge

SEC filings (10-K and 10-Q) contain critical risk information buried in dense legal text spanning hundreds of pages. Investors and analysts need to quickly understand risk relationships and dependencies across multiple filings, but manual analysis is time-consuming and prone to missing important connections.

The Solution

Utilized NLP with LangChain's GPT-3.5-turbo model to extract and categorize risk factors from legal documents. Parsed data from the SEC EDGAR database using BeautifulSoup for data extraction. Implemented graph visualization of risk factor relationships using NetworkX and PyVis, rendered within a Streamlit web app for interactive exploration. Ensured secure environment configuration with python-dotenv for managing sensitive information. Designed user experience in Figma.

Impact & Results

Automated risk factor extraction from SEC filings

Reduced analysis time from hours to minutes

Interactive network graphs showing risk relationships

Comparative risk analysis across multiple companies

Recognized at Stanford Law School Hackathon

Technologies

PythonLangChainGPT-3.5StreamlitNetworkXPyVisBeautifulSoupFigmaSEC EDGAR API

Tags

LangChainGPT-3.5StreamlitNetworkXPyVisFigma
Zakaria Kortam | AI Engineer & Product Engineer