The warnings about the financial crisis of 2008 were all there, hidden in plain sight — but only if you had the time or tools to look hard enough. Hedge fund manager Michael Burry, whose story was told in the film The Big Short, combed through thousands of collateralised debt obligations and realised that there was a ticking time bomb sitting under the US real estate market. But too few people had taken the time that Burry had to run this analysis, and the resulting crash had global repercussions. No one actually had a structured way to track and access this information, which lay buried in piles of legal contracts.

Since then, regulators have tried to gain a better oversight of the risks that banks are taking and understand the potential for financial contagion. They, like Burry, now understand just how important it is to know exactly what liabilities exist in your financial contracts. This means going through the text of these legal documents pulling out the relevant pieces of information from the text.

But imagine you’re a global investment bank with a trillion-dollar balance sheet, and you have thousands upon thousands of these contracts. What do you do if the regulator demands that you tell them what exactly is in them, within 24 hours? How could you do this in time? Traditionally, most banks would bring in an army of lawyers. In fact, countless hours have been spent by junior lawyers sifting through piles and piles of documents in a tedious, repetitive task that was not a good use of their (expensive) time. Today, you can just call Eigen Technologies.

Eigen’s flexible, scalable natural language processing (NLP) technology allows any non-technical user to train up their own machine learning (ML) model to extract structured data from any unstructured text.

At Dawn, we’ve been bullish on the opportunity for software businesses — like Collibra, Quantexa and Dataiku — who can help businesses to drive value out of their data. Such is the potential that the global big data and business analytics market, which was valued at $169bn in 2018, is expected to grow to 274bn by 2022.

But most software tools rely on having clean and ordered data to begin with. And in fact, like the liabilities hidden in financial contracts, more than 80% of an enterprise’s data is actually unstructured, locked up in the text of contracts, emails, reports, or other documents. So finding the company that can best help process this text – quickly, accurately and scalably extracting relevant pieces of information – was the logical next step.

Eigen has developed a game-changing solution. Their natural language processing technology works across virtually any document type and use case. Most NLP solutions typically require a lot of documents for training — usually hundreds, if not thousands. That means it takes an equally long time to train them for a new task and using them for tasks where few documents are available is simply not feasible.

Eigen changes that. Companies can get the same level of accuracy as other NLP players with as few as two, or up to fifty, documents. That’s orders of magnitude fewer than competitors. It can get up-and-running quickly, it’s a good fit for small use cases and scalable for use on larger ones. It analyses documents more quickly than the competition and is up to 30 per cent more accurate. Eigen also empowers end-users who can easily build their own ML model for new use cases, even if they are not trained in data science.

Today, Eigen’s technology is being used for tasks such as legal due diligence and contract review, operational automation and reporting, identifying LIBOR exposure, email triaging and processing, and more.

In future, the company wants to enable people to process textual data from any document and on any use case.

When we really got excited was when we spoke to Eigen’s customers. Goldman, the bank that we referred to at the start of this article, chose to work with Eigen after conducting a bake-off of more than 20 different AI and ML providers — including IBM, Kira, and Seal. Eigen is now working with many more leading industry names such as ING, A&O and Hiscox. In fact, the company is working with 25 per cent of the globe’s systemically important banks including three of the six largest US banks.

In our conversations with them, they were effusive about the company and its potential within their organisations. Many of them had become advocates for Eigen, offering referrals and references.

The remarkable resume of Lewis Liu, Eigen’s co-founder and CEO, also impressed us. He was the first Harvard student to do a double major in physics and fine art, which he followed with a PhD in physics at Oxford, before joining McKinsey. He founded and led the Quantitative Finance & Strategies division of Aleron Partners, and then helped Linklaters as a senior strategic advisor to successfully predict which banks would fail the European Central Bank stress tests.

Lewis brings incredibly deep domain knowledge to Eigen, and has built a world-class around him. It’s clear that we’re investing not just in market-leading technology, but in a founder who combines the expertise, ambition and charisma to attract the highest quality team across product, sales and marketing. With 80 scientists, engineers, strategists and creatives based at the London HQ and at Eigen’s New York office, the company can draw on a wealth of talent.

We’re extremely excited, therefore, to be joining Eigen’s journey, co-leading this $37m Series B funding round with Lakestar and joining Temasek and Goldman Sachs. This round brings the total raised by Eigen to $55m following the Series A round in June last year.

From our perspective, what Eigen is doing now is really just the beginning. They’ve grown sixfold over the last year and landed numerous awards, such as the FT Intelligent Business Award for repapering and a place on the 2019 FinTech 50. Going forward, they’re planning to take their technology and move into fields such as healthcare, government, insurance. This latest investment will only help them to both grow their customer base and continue to push the product even further.