Legal tech startup Luminance, backed by the late Mike Lynch, raises $75M

Generative AI is getting better at interpreting dense texts, and this progress has proved to be a blessing for startups that attack one of the most complex sets of texts that is: the law. It is then logical that we have seen a new eruption of activity in the legal technical room outside the back of the progress in AI in the past year or so.
Legal Technical Startup Eudia to fail Only last week $ 105 million; Genie Ai, based in London elevated € 16 million last year; Harvey Law in the US landed a $ 300 million round led by sequoia; And Lawhive raised $ 40 million to go after ‘Main Street’ American lawyers. The newest addition to that list is ShelterWho invoices itself ‘legal’ AI.
Luminance has claimed to be a very accurate interrogation of legal issues and contracts and has collected $ 75 million in a series C finance round led by Point72 private investments. The round is remarkable because it is one of the largest capital increases due to a pure play legal AI company in the UK and Europe. The company says it has collected more than $ 115 million in the last 12 months and a total of $ 165 million.
Luminance was originally developed by academics in Cambridge, Adam Guthrie (founder and head of technical architect) and Dr. ir. Graham Sills (founder and director of AI). It was funded by seed by the late Dr. Mike Lynch, founder of Autonomy, who died last year in a tragic accident.
Luminance uses what it calls an AI system “Panel of Judges” to automate and increase a business approach to contracts and to increase generation, negotiating and post-execution analysis. The startup uses a patented large language model (LLM) to provide its most important product, Lumi GoWith which customers can send concept agreements to a counterparty and automatically negotiate the AI on their behalf.
Instead of using a GPT (generative pre -trained transformer), Luminance uses what it describes as an LPT (statutory pre -trained transformer) that has been trained at more than 150 million verified legal documents. Many of these documents are not publicly available, which, the company says, makes its platform relatively defensible. Other legal technical startups tend to build on existing LLMs for general purposes.
“It is a domain-specialized AI that is built with lawyers in mind […] They must understand that the outputs are validated and can be trusted, and that is exactly what our specialized AI can achieve, ”said Eleanor Lightbody, the CEO of the startup that the founders took over after the Serie A round.
Lightbody explained that the platform was built with the concept that each model is good in various things. “What you want is to have a mixed model approach, where the models can check each other’s ‘homework’ and you can get the most accurate and most transparent answers,” she said.
She claimed that this approach distinguishes the luminance from its competition, because its customers can use his platform throughout the entire life cycle of contract.
Luminance currently has more than 700 customers in more than 70 countries and contains names such as AMD, Hitachi, LG Chem, SiriusXM, Rolls-Royce and Lamborghini. The workforce has tripled in North America after it had opened three offices in San Francisco, Dallas and Toronto and has expanded his American head office in New York.
The series C also saw participation of Forestay Capital, RPS Ventures and Schroders Capital, as well as existing investors March capital, national grid partners and slaughter and May.