:Harvey: Power Users: "People Call Me :Harvey:"

People call me "Harvey" here because Harvey and Javi are pronounced very similar in Spanish, and I was the one in charge of organizing the pilot, evaluating Harvey, and rolling it out. It started as a joke, but I was finally condemned when Maria, our team leader, called me Harvey in the first training session. My colleague Marina and I are in all the betas and pilots, helping the team with new functionalities and use cases. I think I'm definitely a power user.

We have two different kinds of power users. The quantitative ones are the top users in terms of queries and usage, they're power users because they're more familiar with the tool. Then we have the qualitative ones, people who are using Harvey less frequently, but what they're doing has better output, better quality, or they're pushing the frontier of what's possible.

For me, the power user is the person who is constantly finding new use cases, who is making us use AI better because they are pushing Harvey. If you found a sophisticated way of using Harvey one year ago and you're not improving on that, then for me you're not a top user. Out of 150 users in our legal team, we're following about 30 power users, and about six or seven satisfy both definitions. We have a group that participates in all the pilots and testing, called the “Harvey Taskforce”, a group of power users and champions that are leading the change.

There are three traits we see. First, being younger, which may be the converse of having long-established ways of working. Second, being tech-savvy, and having that ease with programming and technology. For example, I was an IP specialist who was part of the foundation of the legal tech department in my previous firm, and I know how to program. Third, having a heavy workload, for obvious reasons, this incentivizes people to master the technology. They're also more likely to be innovative in the way they use it.

Three years ago, we launched a big transformation project across the legal department on technology, knowledge management, and how we communicate that internally. A big part was creating an innovation mindset. We pushed hard on digital training, a foundation that helped enormously when AI arrived.

We also created a platform where people can share use cases they've discovered. When you see people talking about "I tried this with Harvey" or "How would you do that?," that's the innovation culture we've built. And we try to find a "top user" on each team who can spread adoption organically. The best evangelism happens over coffee, not in formal meetings.

We are also very lucky as the company is prioritizing AI implementation, including the creation of a GenAI Centre. We have the support of the IT team and other corporate areas that help us leverage AI such as mapping processes, calculating ROI, and training.

Play. The best way to learn, not only AI, but technology in general, is playing. Something we're trying to change is to help lawyers lose the fear of failure. For lawyers, it's like "if I start something I need to end it well," but I think the best way to learn how to use technology is to play: don't try to get anything specific, just see what you can do.

Harvey is a secure environment, so the worst thing you can lose is time. Sometimes knowing what not to do is more useful than knowing what to do. You need to understand how the technology works at a basic level, and understand the concepts, so you can design workflows accordingly.