Harvey Introduction

Harvey operates by leveraging custom-trained large language models (LLMs) built in collaboration with OpenAI, specifically tailored for legal work. The platform offers four main components: Assistant (for delegating complex tasks in natural language), Research (for rapid research with accurate citations), Vault (for document analysis), and Workflows (orchestrating multiple specialized models for complete legal workflows). When engaged by a firm, Harvey can be trained on the firm's own work products and templates, similar to onboarding a new employee. The system can handle various tasks including contract analysis, due diligence, litigation support, and regulatory compliance, all while maintaining secure data handling and providing inline citations linking to source materials.