Spellbook's AI-first contract infrastructure
Scott Stevenson, CEO of Spellbook, on building Cursor for contracts
Spellbook is moving from a Word copilot into the operating layer that routes contracts through a company. The product direction is clear, intake, triage, workflow, storage, system of record, and background agents that start work before a lawyer opens a document. That is a different lane from chat based legal assistants, because the moat becomes workflow ownership, contract data, and daily embedded use across legal, procurement, and sales teams.
-
The near term roadmap looks like an AI rebuilt version of CLM, but without the manual setup burden that made older CLM tools underdeliver. Spellbook is targeting teams processing up to 100,000 contracts a year, with automated intake, Slack and email ingestion, and proactive first pass review.
-
Spellbook’s wedge starts inside Microsoft Word, where lawyers already redline contracts, then expands outward into the full contract flow. That bottom up path is especially strong in in house teams, which now drive about 60% of revenue and are growing faster because speed gains map directly to business throughput.
-
Compared with Harvey and Legora, which are closer to broad legal copilots sold into lawyers, Spellbook is specializing around the contract workflow itself. Compared with Ironclad and Icertis, it is aiming to rebuild contract infrastructure with AI first, rather than layering AI onto older workflow systems built around manual fields and configuration.
If Spellbook executes, it can graduate from a drafting tool to the system that companies use to move every contract from request to signature to ongoing monitoring. The next step after workflow ownership is always on agents that watch inboxes, track obligations, surface risk, and quietly push contract work forward across the business.