Federated Learning for Privacy-Focused NLP
Project Overview
A federated learning pipeline trains models on local legal documents, sharing encrypted updates to preserve confidentiality.
About This Project
Federated NLP system to detect risky legal clauses without centralizing confidential contracts.
Tech Stack
Team
Mentors
- Abhimanyu
- Chaitanya Menon
- Shreya
Mentees
- Dhruva
- Emerin
- Tanmay
- Rohith
Methodology
The project follows a structured implementation approach that includes Local document preprocessing and clause extraction, Transformer embeddings with legal-domain models, Federated training with encrypted model updates, and Flower/PySyft style privacy-preserving orchestration. These steps are executed iteratively to validate assumptions, improve performance, and ensure reliable delivery of the final solution.
Expected Outcome
By the end of this project, the team is expected to deliver Cross-client learning without data sharing, and Compliance-friendly risk analysis of contracts. Together, these outcomes reflect both technical feasibility and practical value for demos, evaluation, and future scaling.