Goldman Sachs is backing a London startup using AI to read complex legal documents
- Eigen Technologies has raised £13 million from Goldman Sachs and Temasek.
- The London startup uses AI to read and analyse complex contracts and documents in fields such as law and finance.
- Goldman already uses Eigen, as do Linklaters, ING, and Evercore.
LONDON - Artificial intelligence startup Eigen Technologies has raised £13 million from Goldman Sachs Principal Strategic Investments and Singapore's state investment fund Temasek.
London-based Eigen uses artificial intelligence technology to read legal and financial documents, making it easier for lawyers and bankers to analyse complex contracts - everything from derivatives to land deeds - and find specific clauses. Goldman Sachs is a customer, as are Linklaters, Evercore, and ING.
Eigen's cofounder and CEO Dr. Lewis Z. Liu, who was featured on Business Insider's UK Fintech 35 under 35, said in a statement: "Three and a half years ago we set out to be a truly unique AI company, one that allows our clients to harness the power of their qualitative data to make better decisions. This Series A round underlines our ambitions and is the next major step in our expansion plan.
"With our partners, each of whom is a leader in its sector, we will continue to expand across multiple markets and geographies, as well as doubling down on our investment in research and development, which will always be at the heart of our company."
Eigen already has offices in London and New York, and plans to use the fresh funds to grow these bases and fuel expansion across Europe and into Asia.
The use of AI in document analysis has become a hot area of tech in recent years, with companies such as Luminance and Ravn winning business with major law firms and financial companies. The paperwork-heavy nature of these industries means that AI has the potential to hugely reduce the man hours spent combing through contracts.
While it is not without rivals Eigen claims its technology is simpler and quicker, needing just 30 minutes and a comparatively smaller data set to be trained on in some cases.