Development of RAG with LLM for Enhanced Information Access

In this project, I designed and implemented two RAG-based agents tailored for a corporate environment, leading the development end-to-end:

Legal Contract Explorer: This agent leverages the company’s contract database to provide the legal team with rapid and practical access to key contract details. It enables seamless exploration and retrieval of contract clauses, terms, and conditions, improving efficiency in contract review and analysis.

Operational Knowledge Assistant: This agent utilizes the company's repository of operational documents to assist employees in accessing critical information. It provides step-by-step guidance for operational processes, machine operation manuals, required safety equipment, and more, ensuring a fast and organized information flow across departments.

My responsibilities encompassed every stage of the process, including:

Extracting and cleaning data from diverse sources.

Aggregating data with relevant metadata for better context.

Vectorizing the data to facilitate efficient search and retrieval.

Developing the chain and performing advanced prompt engineering to ensure high-quality results.

Deploying the models via secure endpoints for seamless integration.

Developing and integrating a chatbot within Microsoft Teams, enabling intuitive and direct access to the agents for employees.

This comprehensive approach significantly enhanced productivity, allowing the legal team and employees across departments to access information efficiently and effectively.