Microsoft : and Walter Kluwer Legal & Regulatory partner to explore AI-driven legal workflows

When it comes to productivity, few sectors know better than the legal industry how time-consuming research can be. A daunting amount of information is needed for success and getting the details right is critical. Indeed, a top pain point for lawyers is coping with increased volume and complexity of information, according to a 2021 Future Ready Lawyer study from Wolters Kluwer, a thought leader in the legal space and a global provider of professional information, software solutions and services.

And it's not just lawyers who face this challenge. Knowledge workers could save between four and six hours on average per week if they didn't need to search for or recreate information, according to a 2021 Ziff Davis study on Knowledge Sharing sponsored by Microsoft. That represents 11%-14% of their work time, or over a month of lost productivity per employee per year.

That's why Wolters Kluwer Legal & Regulatory and Microsoft's Modern Work Customer Co-Innovation team (MWCCI) partnered in 2021 to explore potential solutions to legal productivity challenges.

"Wolters Kluwer Legal & Regulatory was the perfect partner because of its deep experience in the legal productivity space, and its desire and ability to do something truly innovative," says Harald Becker, director of Customer Engagements, MWCCI.

MWCCI's co-innovation model involves forming a virtual multidisciplinary product team spanning companies. In this case, it brought together engineers, designers, researchers and others from six countries and three continents. The team works together to thoroughly understand a given problem, define a potential solution, build a working prototype and test it with users-allin an aggressive timeline of about six months. The process is optimized to move quickly and generate mutual learning.

The Wolters Kluwer-MWCCI team began with a deep dive into each other's research, products and platform capabilities. It identified the target user as an attorney in a law firm or corporate legal department and explored multiple possibilities for how to improve the user's workflow.

"We thought that if we could leverage Wolters Kluwer's deep domain knowledge to determine the legal context of the document a lawyer is working on, we could proactively surface useful information that would accelerate the lawyer's workflow and help them produce higher-quality work," says David Jones, MWCCI's principal program manager for this engagement. "This could include analytics for legal arguments, risk analysis for contracts, subject matter experts within the firm-oreven just the right legal research or template."

In just a few weeks, the team articulated the concept it wanted to build: a tool that could bring legal professionals relevant, high-value content from a variety of data sources within their existing workflows, which, for many lawyers, means within Microsoft Word.

"We started to think of the vision for our concept as a kind of legal coach," says Peter Backx, vice president of Product Management, Wolters Kluwer Legal & Regulatory. "We believed that it could become something that helps lawyers improve the outcome of their work."

It was clear early on that the team would need to leverage machine learning, specifically document understanding AI-theability for machines to read and understand the content of documents. This raised questions about users' confidence in the information that the algorithms were bringing them.

"We knew users might wonder, 'Why is the system showing me this?' 'Is this the right information?' 'Has anything been overlooked?' 'What determines what's important to display and what isn't?'" Jones explains. The team agreed that intentionally building a foundation of trust for the system would be critical. Designers explored different ways of explaining within the interface why a given resource was suggested, which they believed would support users' trust in the system.

Within a few months, the team began to build. It split into two parallel tracks, one focused on user experience and visual design, the other on technical architecture. The UX design track generated multiple potential user-interface designs. "It was fun to debate the merits of each and reach a consensus on the approach that would best address the customer's needs," says Magdalena Sowula, lead product manager, Wolters Kluwer Legal & Regulatory.

The technical track was more of a challenge. The team tried multiple ways of extracting legal context from documents but failed, at first, to get useful results. "For a period of time it looked like our project might fail," Jones says candidly.