Harvey bundles software with ex-lawyer services

Harvey is selling adoption, not just automation. In legal AI, the hard part is not getting a model to answer a question, it is getting lawyers to trust it inside live drafting and review workflows, use it often enough to change habits, and renew at enterprise prices. That is why Harvey pairs software with ex-lawyers who help configure workflows, onboard teams, and push usage above the renewal line.

• The service layer looks a lot like vertical AI peers. Hebbia also uses ex-bankers and lawyers as an in house engagement team to build templates, wire data sources, and manage operational change, because enterprise AI value comes from fitting into a real workflow, not from dropping in a chatbot.
  1 sacra 2 sacra
• The workflow itself is concrete. Harvey helps with contract drafting, review, and research, and later added integrations into systems like iManage, NetDocuments, LexisNexis, and Word. That moves it from a side tab for occasional questions into the software lawyers already use to edit documents and look up precedent.
  3 sacra 4 sacra
• This matters more now that model quality is less proprietary. As frontier models caught up on legal reasoning, Harvey shifted away from selling a special legal model and toward selling pre-configured agents, domain setup, and customer specific implementation. The moat moved from model weights to workflow, trust, and usage data.
  1 sacra 3 sacra 4 sacra

Over time, this bundled services model should become more productized. The near term winner in legal AI is likely the company that can turn repeated onboarding work into reusable templates, agent builders, and integrations, so each new firm needs less handholding while the product still feels tailored to that firm’s way of practicing law.