Case studies
80% Accuracy Boost! See How Codvo Revolutionizes Legal Reviews with AI
.png)
The client is a Legal Tech platform that specializes in providing innovative solutions to streamline the contract review process for legal professionals and organizations.
They are committed to making the process of contract review more efficient, accurate, and cost-effective by leveraging cutting-edge technology.
The client approached Codvo.ai to develop an accurate and reliable question-and-answer tool for legal contract review. They wanted a tool that could accurately answer complex questions about legal clauses with a high level of precision and recall. The tool was expected to significantly reduce the time and effort required for contract review.
The business challenges faced by the client were significant. Legal contract review is a complex process that requires a high level of expertise and attention to detail. The process is time-consuming and can be error-prone, leading to delays, inaccuracies, and additional costs. The client wanted to develop a tool that could automate the contract review process, reduce the time and effort required, and maintain accuracy and reliability. The challenge was to develop a tool that could accurately answer complex questions about legal clauses with a high level of precision and recall.
.png)
Codvo.ai adopted a data-driven approach to develop the question-and-answer tool for legal contract review. Various large language models, such as RoBERTa, DeBERTa, and LegalBERT, were fine-tuned for legal contract review on CUAD datasets. RoBERTa was shortlisted based on performance. Pre-processing techniques, such as tokenization and embedding, were used to prepare the data for the models. The CUAD dataset was used to train and evaluate the models, which include over 500 legal contracts with over 60,000 sentences.
The LLM (Large Language Model) was developed using the fine-tuned RoBERTa model and was used to identify 41 key entities from any form of the contract documents, such as party names, addresses, indemnities, and more. The LLM model was integrated into the AWS service, enabling clients to easily access the answers to each question within any document.
The tech stack used by Codvo.ai for this project included AWS S3, Lambda, Sagemaker, TensorFlow, Hugging Face, and PyTorch. These technologies were chosen for their scalability, reliability, and ability to handle large volumes of data.
.png)
.png)
.svg)