AL Interview: Tim Pullan, UK Legal AI Co. ThoughtRiver

Tim Pullan, the CEO of UK-based legal AI company, ThoughtRiver, came to the world of legal tech start-ups through a series of steps.

Having started professional life as a trainee at insurance law firm Kennedys, Pullan moved through several legal roles, including partnership at the then Lawrence Graham, which is now part of Gowling WLG, then a stint in Asia for consumer credit data company Experian, then joining the Singapore office of Taylor Vinters as Head of Technology, before returning to Cambridge.

During this journey through different roles and firms, Pullan became increasingly focused on how technology could be applied to legal contracts in order to remove risk, much as Experian focuses on gathering data to assess risk on personal credit-worthiness.

The result was the creation of ThoughtRiver (TR) in January 2015, a joint venture with his current firm, Taylor Vinters.

In a nutshell, the central plank of TR’s offering is using natural language processing (NLP) to scan and assess risk levels in elements of a legal document.

The system has been trained to identify certain legal clauses and how they would normally appear, as well as what certain types of legal document should look like and include. When the system spots something amiss, or that may need a second look, a dashboard provides a visual risk assessment, allowing the lawyer, whether inside a law firm or inhouse team, to quickly check and respond to the flagged issue.

The benefit is that it greatly speeds up the process of document review and helps corporates to eliminate risk in an efficient and economical way.

One other key aspect to the offering is that, as Pullan says, corporates can shy away from doing a thorough risk assessment on lower value contracts because the cost/benefit formula normally doesn’t reward lots of manual legal labour in these matters. Using TR’s AI capability allows a company to risk assess large numbers of lower value contracts and find very quickly where there may be outliers that need a second look by a human lawyer.

And that is where TR has been for a while, as it steadily conducted Beta testing with a number of corporates. However, 2017 has been a year of significant milestones and will see more changes to come. Artificial Lawyer caught up with Pullan to hear why 2017 may well be TR’s break out year.

The first fact is that in May this year the company received $1m in investment from a group of investors that included former Vodafone UK CEO Guy Laurence and Michael Findlay, chairman of Morgan Sindall and former co-head of investment banking for Bank of America Merrill Lynch, which certainly helped and showed a high degree of confidence in TR as a going concern and not just as a ‘nice idea’.

Tim Pullan, ThoughtRiver

‘We have scaled up the tech team, which was small before and that is where the bulk of the money is being used. We’ve also invested in our sales team and also in the contract intelligence team,’ says Pullan.

‘We need lawyers to deal with the configuration of ThoughtRiver,’ he adds, as the AI needs to be trained what to look for in all that unstructured data of so many different types of contract. And, that’s a job you need lawyers for, as who else will be able to really understand the complexities of what to look for?

There will also be more of a push on getting lawyers in the team to talk to clients to learn about what they really want TR to examine and surface in terms of potential risk, or in terms of other useful data that can be found in the mass of contracts TR is working through.

Pullan adds that over the next three months the company’s staff will be running a development programme with selected clients to address specific contract risk problems, for example in the area of GDPR data regulation.

‘In the past I did many different jobs,’ Pullan explains, having experienced the joy and challenges of being a founder who at first had no choice but to do a lot of key tasks themselves. However, Pullan notes he will remain CEO and the key decision-maker about where TR is heading.

And where TR is heading is important. The first item to note is that this latest funding round is just the beginning. A much larger round is expected later this year which will be a multiple of the May investment. That will allow even further scale increases and more clients to be served.

The other important step forward is what Pullan describes as making better use of the insight they have gathered in their work analysing, for example, sales contracts that a corporate client may wish to use TR to risk assess.

Pullan says he sees this capability as a little different to other CATs systems (or Contracting Automation Tools) as TR’s approach would be more flexible. Rather than working off what are often a narrow set of templates with clear red lines as to how a contract can be formed, TR would be focused on a broad range of potential documents. However, the end goal would be the same: to capture that extracted data and turn it to good use for the client.

For example, TR could build up a picture for the client of what are the most risky terms in certain types of contract over a long period of time, which are the parts of the contract that lead to later problems with suppliers and how are they phrased and constructed? How much do those issues cost on average to resolve? Are there certain parts of the company triggering higher levels of risk in the agreements they make compared to others…..and what terms are they adding/excluding in their contracts that is making this happen?

Once you begin to see legal documents as data and specifically as data that can be cut in dozens of ways to provide insight just as one can with a heavily populated Excel sheet, then TR and other legal AI data analysis systems take on a whole new meaning. What you can find then becomes a question of what you can imagine looking for.

Pullan therefore sees a bright future for his creation and multiple use cases for it. However, at the heart of the system are themes that have been steadily developing in the world of legal AI for some time: efficiency, the removal of low value work from lawyers’/paralegals’ hands and better use of unstructured legal data.

‘We want to emancipate lawyers from the dross, to allow them to be the advisers they always wanted to be,’ Pullan concludes.

And that sounds like a very positive company motto to have.