All Projects
AI Systems1 September 2024

AI Corporate Governance Platform

Annual reports, nomination committee documents, and proxy statements fed into an AI pipeline that generates complete governance, remuneration, and ESG reports automatically.

PythonOpenAIRAGFastAPIReact

Outcome

Saves 5 hours per report. Governance review cycle cut by ~70% for a team of investment analysts.

Live Demo

Live Pipeline
📥

Document Ingestion

🔗

Chunking & Embedding

🧠

AI Analysis

Report Generation

Complete

$ awaiting pipeline start...

The Challenge

Investment analysts at The Investment Association were spending 4-5 hours per company manually reviewing annual reports, nomination committee disclosures, and proxy statements to extract governance and remuneration data. The process was inconsistent, bottlenecked by individual availability, and produced outputs in varying formats that were difficult to compare across the portfolio.

How It Was Built

Built a multi-stage RAG pipeline that ingests PDF governance documents, chunks and embeds them using OpenAI embeddings, then queries the vector store with structured prompts targeting board composition, remuneration ratios, ESG targets, and audit committee independence. A FastAPI backend orchestrates the pipeline and a React interface allows analysts to trigger processing, review extracted signals, and download formatted output reports in three categories: corporate governance, remuneration, and ESG. The system applies IA-standard templates to ensure consistency.

The Result

Processing time dropped from 5 hours to under 10 minutes per company. Analysts can now run the pipeline overnight across a full portfolio and arrive to reviewed outputs the next morning. The consistency of extraction improved materially because every company is queried against the same structured prompts rather than relying on individual analyst interpretation.

Technology Stack

PythonOpenAIRAGFastAPIReact