Harvey is an AI platform designed from the ground up for the practice of law rather than adapted from a general-purpose chatbot. Its core surfaces include Agents for orchestrating multi-step legal tasks, Vault for storing and analyzing large document sets, Knowledge for legal, regulatory, and tax research, Spaces for shared client and co-counsel workspaces, Command Center for adoption analytics, and Memory for retaining user preferences across sessions. Together these pieces aim to let a firm or legal department run matters, research, and review from one place instead of stitching together disconnected tools.
The product has moved toward an agentic model, where complex matters are decomposed and routed to specialized agents that produce drafts, chronologies, and fact patterns for lawyer review. Workflow coverage now spans contract review and negotiation, regulatory horizon scanning, legal intake, and litigation support such as facts and chronologies. Connectors and an MCP-based library let teams bring Harvey into the tools they already use.
Commercially, Harvey serves thousands of legal organizations and has raised significant venture funding, positioning it as one of the best-capitalized entrants in legal AI. Access is arranged through a demo and enterprise agreement, so exact seat and usage pricing is quoted per organization rather than published as a public list price. Security, ethical walls, and data governance are treated as first-class requirements given the sensitivity of the material involved.