While some firms have created subsidiaries to develop and sell their proprietary legal tech, other firms have taken a different route, recruiting software developers in-house to create technology that exclusively benefits the firm and its clients.

Fish & Richardson joined the latter group when it formally spun off its legal technology solutions group from its IT department and promoted applications, development and support director Beau Mersereau to head the new group.

The group’s team of application specialists, business analysts, quality assurance automation engineers and developers that report to Mersereau are tasked with creating client-facing and internal legal tech solutions.

In a conversation with Legaltech News, Mersereau discussed how the group manages the law firm’s massive troves of data, the opportunities in leveraging machine learning, and the reasoning for moving the group out of the IT department.  These answers have been edited for clarity and brevity.

Legaltech News: Why was the legal technology solutions group spun off from Fish & Richardson’s IT department?

Beau Mersereau: We wanted to focus primarily on new technologies and innovative services, and bifurcating it allowed us to focus on the areas that were important to our clients. That also allowed the IT department to focus on technology and providing better services to our firm with less distractions.

What are some of the new solutions the group is developing? 

We are piloting machine learning to auto-classify documents or incoming mail from the Patent and Trademark Office that will allow us to route mail automatically to the appropriate teams. Eventually, we hope to start doing other things from auto-classifying time cards to having better data within our pricing group.

We hired a data scientist this year and he’s been helping us a lot. He used to be an astrophysicist, and it’s pretty interesting to have someone from a different field looking at our data and trying to understand it. Law firms have a lot of data and it’s not always easy to find the nuggets of good information in there.

What are some of the challenges or benefits of having that trove of data?

A lot of other firms have their knowledge, information and data locked into documents. We’ve been using document assembly since 2001 and because of that, especially on the patent and trademark side, we actually store all the information we use to generate our documents in a database. We have a lot of information that we track and it’s not always easy to look for patterns and find out, for example, how long it takes to do something before the Patent and Trademark Office. But we have that data now and we are learning from it.

How important is machine learning-powered legal tech?

Law firms are constantly looking for ways to become efficient and our clients are demanding we be more efficient and innovative [with] our use of technology, and machine learning is a game changer. When you pair that with the computing power that you have with the cloud, which is far more than what we have on premises with our servers, we have the ability to build models more quickly to help us classify information.