All Insights

Retrieval-Augmented Generation (RAG) in Legal Technologies

Published on
View Source Code

Abstract

The application of Large Language Models (LLMs) in legal technology presents significant risks regarding hallucinated citations. This paper explores the efficacy of Retrieval-Augmented Generation (RAG) architectures in mitigating these risks.

Methodology

We indexed 10,000 public court rulings using OpenAI text-embedding-3-large and implemented a hybrid search approach (Dense Vector + BM25).

Results

  • Accuracy IncreaseHybrid search improved retrieval accuracy by 22% over pure vector search.
  • Hallucination ReductionStrict prompting combined with accurate context retrieval reduced hallucination rates to <0.5%.

    • Research Gallery

    Retrieval-Augmented Generation (RAG) in Legal Technologies gallery image 1