How AI and machine learning can (and can’t) streamline time capture for law firms

Today’s legal clients are savvier than in decades past, demanding better reporting and greater levels of intelligence from their law firms. They want more insight into what they’re paying for, and they expect higher predictability on the outcomes. To provide exceptional service, law firms are expected to provide regular bills that offer an in-depth view of how staff time has been spent on clients’ matters.

Consequently, firms have turned to advancements in time capture technology to free lawyers from tedious administrative timekeeping tasks. Many law firms are hopeful that AI will be the silver bullet — but although AI can be extremely helpful with time capture, it’s not a universal remedy.

Dreaming of do-it-all technology

Most lawyers wish for a solution that can perform their most laborious tasks for them, including tracking hours. Today’s automated time capture tools do indeed perform well at helping lawyers with timekeeping, but the notion that AI and machine learning can handle all timekeeping nuances for them remains a dream — at least for now.

AI and machine learning are, by now, proven time capture technologies. As a lawyer works on a matter throughout the day, AI-based time capture can passively identify and categorize a daily diary, simplifying task re-creation and time logging. AI quickly connects important information — such as phone numbers, emails, and other important details — to help link activities to certain clients or matter IDs. As the lawyer continues to work, a well-trained AI can start associating future activities with their corresponding clients and matters. This simple use of AI can save lawyers hours of effort, resulting in more time spent on high-value, client-related activities instead of administrative tasks.

Training the AI engine

Simple tasks can be easy for AI to understand, but more complex tasks can require additional training. AI what’s called “emergent behavior,” meaning it doesn’t solve problems or approach a solution the exact same way that a human brain does. As a result, you may get different results than you were expecting, and you may need to continue “training” the technology to get an optimal result.

Emergent behavior in AI can be a good thing when you want a machine to look at a problem in a different way than your humans have already tried. But lawyers (and clients) require elements of time capture, like narratives, to be very precise, just as a lawyer would provide them — which requires a very sophisticated level of AI, and a lot of time and training. Of course, the time capture narrative is one of the most important elements of the timesheet, as far as a client is concerned: It’s how they get their detailed understanding of what their law firm is spending their money to accomplish.

Understanding AI’s limitations  

Software companies invest significant effort in training their AI engines. But since every law firm has its own unique requirements — requirements that even a well-trained AI engine cannot recreate out of the box — additional firm-specific training must take place. Then, over time, the engine will gain additional information through hours of lawyers’ and staff’s inputs — information that helps refine its algorithms.

Still, AI can struggle to imitate the unique ways that some lawyers like to create time-entry narratives. It must analyze a significant number of narratives from an individual lawyer to become smart enough to create usable entries for that specific lawyer. AI solutions measure the confidence, or accuracy, of their predictions — and as the engine continues to learn, accuracy will rise, eventually reaching confidences score into the 90s. That’s excellent progress, but still leaves room for improvement.

Making lawyers’ lives easier with technology that’s ready today

For the last two decades, Intapp has focused on making lawyers’ lives easier. Today, AI is a big part of the Intapp OnePlace platform, although many users don’t even know that it’s AI that’s assisting them in their everyday work.

Across the entire platform, AI has been embedded into the Intapp OnePlace workstreams in ways that minimize change and disruption, and don’t introduce errors or create additional work — an important goal in our design. For example, in Intapp Time, AI can identify a document ID and associate it with a lawyer’s timecard, offering a helpful “nudge” to the lawyer. The process is frictionless to users, and creates no incorrect narratives to edit or delete — and that’s exactly the experience it should be.

AI is a powerful technology, but it needs to be used correctly with lawyers’ timekeeping. Lawyers may be sad to learn AI will not be completing their timecards today, but in the meantime, let Intapp Time and the Intapp OnePlace platform use AI to help “nudge” your lawyers to be more efficient while making your firm more profitable.

Put the power of AI-assisted time entry to work for your firm: Schedule an Intapp Time demo.