Bitesize InsurTech: Eigen Technologies
August 23, 2018 Greg Brown
Eigen Technologies uses natural language processing (NLP) to automate the extraction and classification of information from documents using machine learning.
The business shot to prominence in the insurance industry this month with its announcement of a partnership with Hiscox.
The business was founded in 2014 by Lewis Liu after his PhD in physics at Oxford University (where he was the lead inventor of a “new class of polarization-controlled high harmonic X-ray lasers” – explanations on a postcard, please) and Jonathan Feuer, a Managing Partner of CVC Capital.
Eigen’s technology allows for the automated extraction of data from diverse types of documents at scale. In the words of Tom Cahn of Eigen, this solves the ‘qualitative data problem’, referring to data that is not numeric and is therefore hard to process. He adds that 98% of the world’s data has never been analysed. “CFOs can get you the numbers from Oracle, but the General Counsel can’t work out exposure on a policy-by-policy basis.”
One of the business’s early breaks occurred in the banking industry in automating compliance with US financial regulation. Eigen’s technology has enabled its clients to automate regulatory processes at scale. Referring to one particularly big project at a US investment bank, Tom said: “many similar institutions called in armies of lawyers, but this one decided to try technology-led options.” Eigen beat the field in “exam conditions” and won the contract.
Tom claims that Eigen stands out in several ways. First, clients retain their data – it is never used to power models for other companies. Second, Eigen’s platform allows non-technical users to control the machine learning process from start to finish. Third, clients can start training the algorithm with just a few dozen documents – although it is important to remember that these documents need to have a level of standardisation to allow the software to build an analysis model. Finally, Eigen’s technology can be deployed either in the cloud, or on premise inside a client’s internal systems.
Eigen raised its £13m Series A round in June 2018. The round was led by Goldman Sachs PSI and Temasek, the Singaporean investment fund. Clients include Goldman Sachs, Evercore, Linklaters, Hiscox and ING.
The Oxbow Partners view
It’s easy to get excited about Distribution InsurTechs and describe their unique features. It’s harder to express similar feelings about a Supplier InsurTech without getting into the technical detail – which we won’t.
In some ways, Eigen is experiencing a ‘perfect storm’ in the industry. On the one hand, insurers are focused on cost, which gives Eigen its efficiency story. On the other, insurers are trying to collect and use more data for anything from improved customer experiences to better underwriting processes. Given that much of this data is buried in lever arch files, Eigen has a role to play here too.
The challenge for the industry will be designing efficient implementation paths. It is important for incumbents not only to focus on digitising information, but to be clear on its purpose. In other words, it’s important not to digitise for digitisation’s sake, but to see digitisation as a tool in a general march towards better customer experience, greater process efficiency or improved use of data. (Oxbow Partners helps clients in all three of these areas – drop us a line if you’d like to hear more.)
One thought we have is on the relative speeds of NLP-enabled processing compared to the speed of general technological development. In theory, as the industry trades more digitally, the volume of unstructured information will go down, or will be confined to increasingly complex or bespoke items like reinsurance contracts. There’s no shortage of use cases for Eigen Technologies (and companies like it – see our our bitesize on Impact 25 Member Risk Genius) in the market now, but they’ll need to keep ahead of general digitisation efforts.