Predictive analytics pioneer CourtQuant is in the process of being wound up, with its chief executive blaming cautious lawyers, a lack of data and the impact of the pandemic.
Jozef Maruscak said lawyers were reluctant to “rely completely” on the technology unless they could explain how the algorithm worked, and there was a lack of data to train it.
A further problem was the pandemic, which made it harder for CourtQuant to raise the money it needed for the next three to five years.
Mr Maruscak, 26, a former Cambridge law student, said: “We built a good product in terms of efficiency, but law firms asked for explainability – how to explain what was happening behind the predictive analytics.
“Lawyers are reluctant to rely on a prediction when they don’t know where it is coming from or how the model is making its predictions.”
Mr Maruscak said that to explain required “significant resources” and enough data, but the data situation in the UK was “not good”.
He accepted there was a danger of bias in predictive analytics, as there was in any artificial intelligence model, but said that if there was enough data, you could identify the bias and get rid of it.
After lockdown, the team at CourtQuant met to decide if the company could go forward. They had “offers on the table” in terms of investment, but “they were not good enough to persuade the team that CourtQuant could become a successful business long-term and that technical challenges can be confidently solved”. There was a “general consensus” by May that the company should close.
Mr Maruscak said the company had not been officially wound up, but this would happen in the coming months.
It had been named in Deloitte Legal’s first lawtech start-up incubator last December, but Mr Maruscak said Covid-19 meant little actually happened.
CourtQuant was created in 2018 after Mr Maruscak and the other two Cambridge law graduates behind CaseCrunch went their separate ways.
He said there was a “huge opportunity” for the use of predictive analytics in smaller-scale litigation funding, which is what CourtQuant was working on for its client Apex Litigation Finance.
Apex, which specialises in funding for small to medium-sized cases, announced earlier this month that it was developing its own in-house predictive analytics system following its “successful relationship” with CourtQuant, building on this “experience and expertise”.
Mr Maruscak said it was too expensive to get counsel’s opinion on litigation funding cases worth from £10,000 to £100,000, but funders needed a way to “smartly decide” what to do.
He said it was a “natural step” for Apex to find a way of building a predictive analytics system in-house.
CourtQuant also worked with a debt purchasing company wanting to know which debts to buy, and a “large” European after-the-event insurer wanting to know which cases it would win.
“I think predictive analytics will have a good future, but maybe we were in the market a little too early,” Mr Maruscak said.
“I don’t want to be negative about lawyers. They’re dealing with very important matters I understand why they are reluctant to rely completely on technology. That view will change over time.”
He said predictive analytics would “bring transparency into the law” and clients would be happier if they understood more about whether they were going to win or lose.
He is now working as head of business development for Sudolabs, a technology company based in central Europe.