The Tokyo-born legal AI company teaching machines to read contracts the way a lawyer would - clause by clause, from your side of the deal.
Every deal has a bottleneck, and often it is the legal review. A contract lands in an inbox, a lawyer opens it in Microsoft Word, and the slow work begins - reading each clause, comparing it against the standards the company will accept, marking what is missing and what is dangerous. LegalOn Technologies was built by two people who did that job themselves and grew tired of catching the same mistakes by hand.
Founded in Tokyo in 2017 by former corporate attorneys Nozomu Tsunoda and Masataka Ogasawara - both alumni of the Japanese firm Mori Hamada & Matsumoto - the company set out to encode legal judgment into software. Its flagship product, Review, reads an uploaded contract clause by clause, flags risks, and suggests edits based on playbooks written by lawyers and tuned to each customer's legal positions. The company says it cuts review time by up to 85%.
That combination - large language models steered by attorney-authored playbooks - is the heart of LegalOn's pitch. The AI does not freelance. It works from standards a human lawyer approved, and it shows its reasoning, which is what a legal team needs before it will trust software with a binding agreement.
The approach worked first at home. In Japan, LegalOn became close to ubiquitous, reaching roughly a quarter of the country's public companies and, by its own count, 87% of Japan's Fortune 500. On the strength of that, it raised more than $200 million, expanded to the United States and United Kingdom, and in July 2025 closed a $50 million Series E led by Goldman Sachs Growth Equity.
It also did something notable for a vertical AI company: it partnered with OpenAI. The non-equity deal gives LegalOn access to OpenAI's most advanced models and pairs engineers from both companies to build legal AI agents - a bet that domain expertise plus frontier models beats either one alone.
In-house counsel spend a large share of their time on contracts that look almost identical to the last hundred they reviewed - NDAs, MSAs, vendor agreements. The work is slow, repetitive, and unforgiving: one missed clause can cost real money. It is also inconsistent, because a tired reviewer on a Friday afternoon may not flag what they would on a Monday morning.
LegalOn's answer is to remove the grunt work without removing the judgment. Its "point of view" analysis reviews a contract from your side of the deal - buyer or seller, licensor or licensee - so the risks it flags actually match your interests. The AI is fast; the playbook keeps it consistent. What lawyers do with the reclaimed hours is the quieter story: less redlining boilerplate, more actual counsel.
AI contract review that analyzes agreements clause by clause, flags risk, and suggests edits against attorney-authored standards.
50+ attorney-written playbooks for NDAs, MSAs and more - plus custom playbooks to encode your own legal positions.
A generative AI assistant for legal questions, drafting help and document analysis inside the legal workflow.
Intake, track and manage legal requests and matters across the in-house team in one place.
AI-powered contract translation across 28 languages for reviewing global agreements.
Brings review and revision directly into Microsoft Word - meeting lawyers where they already work.
LegalOn sits in a fast-crowding field of legal AI - alongside Spellbook, Harvey, Ironclad, Robin AI, Luminance and GC AI. Its differentiators are focus and trust: it is contract-first, lives inside Word, and anchors every answer to a lawyer-authored playbook. Large enterprises increasingly pair it with a contract-lifecycle system like Ironclad, using that as the backbone and LegalOn for the AI review itself.
Note: LegalOn declined to disclose its Series E valuation. Reported annual revenue is approximately $66-67M.
Ex-Mori Hamada lawyers Nozomu Tsunoda and Masataka Ogasawara found the company (as LegalForce), building on AI research from Kyoto University.
The platform ships in Japan and rapidly signs thousands of companies.
Enters the United States and names Ravel Law co-founder Daniel Lewis as US CEO, backed by a SoftBank-led round.
Unifies under the LegalOn brand globally and rolls out its Word add-in and playbook library.
Adds Assistant, Matter Management and 28-language Translate - broadening into a full legal workflow platform.
Goldman Sachs leads a $50M round as LegalOn partners with OpenAI to build legal AI agents; US/UK business quadruples.
What makes LegalOn unusual is that lawyers are not just customers - they are inside the product. The company employs attorneys to author the playbooks and legal standards its AI relies on, which is why its output reads like guidance rather than guesswork.
Kyoto University Law graduate and former Mori Hamada attorney who co-founded the company after growing frustrated with contract-review errors.
Former corporate lawyer who co-founded the company alongside Tsunoda in 2017.
Stanford Law graduate who founded Ravel Law (acquired by LexisNexis, pioneered Judge Analytics) before leading LegalOn's global push.
It makes AI software for in-house legal teams and law firms, centered on contract review. Its tools read contracts clause by clause, flag risks and suggest edits based on attorney-authored playbooks - primarily inside Microsoft Word.
It was founded in Tokyo in 2017 by former corporate attorneys Nozomu Tsunoda (Group CEO) and Masataka Ogasawara. Daniel Lewis, co-founder of Ravel Law, serves as global/US CEO.
More than $200 million total, including a $50 million Series E in July 2025 led by Goldman Sachs Growth Equity. The company did not disclose its valuation.
It anchors its AI to attorney-authored playbooks and supports "point of view" analysis - reviewing from your side of the deal - emphasizing consistency and trustworthiness over open-ended generation, and it works inside Word.
More than 7,000 organizations worldwide, from enterprises to law firms. It is especially dominant in Japan (about a quarter of public companies) with fast-growing US and UK adoption.