Summary
Luminance raised $75 million in Series C funding led by Point72 Private Investments, bringing total funding to $165 million. The company, founded by Cambridge academics and seed-funded by the late Mike Lynch, plans to accelerate US expansion where 40% of revenue is already generated.
Legal AI pioneer Luminance has raised $75 million in a Series C funding round, marking one of the largest investments in a specialized legal AI firm across the UK and Europe.
The round was led by Point72 Private Investments, with contributions from Forestay Capital, RPS Ventures, and Schroders Capital. Existing investors including March Capital, National Grid Partners, and Slaughter and May also participated.
This brings Luminance's total capital raised to $165 million, with more than $115 million raised in the past year alone (including a $40 million Series B in April 2024).
Founded in 2015 by Cambridge academics Adam Guthrie and Dr. Graham Sills, Luminance uses generative AI to help law firms with contract drafting, review, and management. The company was seed-funded by the late Dr. Mike Lynch, founder of Autonomy.
Key differentiators:
- **Proprietary Legal Pre-trained Transformer (LPT)**: A model specifically designed for legal work
- **Training Data**: Trained on more than 150 million verified legal documents, many not publicly available
- **Language Support**: Understands over 80 languages, ideal for global firms
- **Panel of Judges Architecture**: Uses specialized AI for contract analysis
CEO Eleanor Lightbody stated: "This funding is all about innovation, expansion and scaling. It supercharges our US growth, where 40% of our revenue is already generated, and will fuel key hires and new offices across the US, APAC and Europe."
Current customer base:
- More than 700 clients across over 70 countries
- Notable clients include AMD, Hitachi, LG Chem, SiriusXM, Rolls-Royce, and Lamborghini
The investment reflects strong institutional confidence in specialized legal AI, particularly solutions built from the ground up for legal work rather than adapted from general-purpose models.