Aalap: AI Assistant for Legal & Paralegal Functions in India

TL;DR

Aalap is a fine-tuned Indian legal domain-specific AI assistant built on Mistral 7B, outperforming GPT-3.5-turbo in legal reasoning tasks and aiding legal professionals with day-to-day activities.

Contribution

We developed Aalap, a domain-specific legal AI model for India, focusing on legal reasoning, with performance surpassing GPT-3.5-turbo on legal tasks.

Findings

Aalap outperforms GPT-3.5-turbo on 31% of test data.

Aalap matches GPT-4 performance on 34% of test data.

Training emphasizes legal reasoning over recall.

Abstract

Using proprietary Large Language Models on legal tasks poses challenges due to data privacy issues, domain data heterogeneity, domain knowledge sophistication, and domain objectives uniqueness. We created Aalalp, a fine-tuned Mistral 7B model on instructions data related to specific Indian legal tasks. The performance of Aalap is better than gpt-3.5-turbo in 31\% of our test data and obtains an equivalent score in 34\% of the test data as evaluated by GPT4. Training Aalap mainly focuses on teaching legal reasoning rather than legal recall. Aalap is definitely helpful for the day-to-day activities of lawyers, judges, or anyone working in legal systems.