Project aims to find risky clauses in lengthy legal documents, along with that it generates pdf files according to provided data(given questions) and given templates, Different models are used for different tasks including GPT, zero shot learning and few shot learning models.