Skip to content
Back to portfolio

Enterprise AI

AI Knowledge Registry

An AI-ready knowledge layer built from thousands of internal technical documents.

Tech Intern at Eli Lilly and Company · May 2025 – August 2025

Public descriptions omit internal documents, proprietary metadata, enterprise source identifiers, and retrieval configuration.

Problem

Thousands of internal technical documents needed to be transformed into a structured, searchable, AI-ready knowledge layer.

My contribution

Built Python-based ingestion pipelines with metadata-driven tagging, automated classification, and retrieval workflows to make internal technical documents AI-ready.

Simplified architecture

Simplified: Python-based ingestion pipelines apply metadata-driven tagging and automated classification, feeding retrieval workflows over internal technical documents.

Outcomes

  • Processed approximately 3,200 internal technical documents.
  • Reduced manual search effort by approximately 60 percent.

Technologies

Python