Overview

Due Diligence Pipeline is a solution designed to help users generate comprehensive KYC profiles on individuals from publicly available sources. The project focuses on delivering an automated, efficient, and structured research experience while providing citation-backed insights and profile documentation.

Whether you are a business user, researcher, or general audience member, this project aims to make the process easier and more accessible.

Current Version

Key Features

Feature 1 is Automated Multi-Phase Research

The pipeline runs 16+ parallel web searches per profile, validates sources for relevance, and extracts structured data using a local LLM, covering shallow attributes and deep content sections simultaneously.

Feature 2 is AI-Powered Section Writing

Profile sections are written by a combination of a local language model and OpenAI GPT-4o — ensuring zero cross-section repetition.

Feature 3 is Structured Word Document Output

Each profile is generated as a formatted Word document following a designed template, with inline citations linked to their source URLs, shallow attribute fields populated from Phase 1 dedicated searches, and a photo retrieved automatically.

Use Cases

This project can be used for:

• Supporting operational workflows
• Improving decision-making processes
• Automating repetitive tasks
• Enhancing user productivity
• Providing visibility into important information

For Organisations

• Standardises profile format and quality across all analysts
• Scales to any number of individuals without a proportional increase in research effort
• Maintains a clear audit trail through inline source citations on every fact