How Ballard Spahr Modernized Analytics from Data Warehouses to Lakehouses
How do you turn legal data into a modern platform that powers analytics, self-service, and AI across one of the nation’s most innovative law firms?
In this episode, Lisa Mayo Haynes, Senior Director of Technology Innovation, and Ian Lopez, Manager of Technology and Innovative Solutions at Ballard Spahr LLP, share how the firm modernized its data warehouse into a lakehouse that unites structured and unstructured data: powering real-time dashboards, predictive insights, and generative AI tools like Ask Ellis and Ballard X-Ray. They outline how Ballard Spahr built a trusted data foundation that drives adoption and measurable impact through accurate, effectively-dated data, secure access, and a shift from reactive to predictive analytics.
Lisa and Ian also discuss how their team built data confidence across the firm: teaching data literacy, simplifying access, and scaling early wins into lasting transformation. From major time savings to AI tools that free attorneys for higher-value work, they show how Ballard Spahr is turning information into a future-ready, trusted data ecosystem.
-
Description text goes hereTranscription
ENTEGRATA | OVERRULED BY DATA | LISA MAYO-HAYNES
Episode Transcript
This has been generated by AI and optimized by a human.
[00:00:00] Lisa Mayo Haynes: If you don't have the trust of your firm with accurate data, then you're really cooked. Good quality data is the foundation of any strategy, an IT strategy, an innovation strategy. You're just not going to be able to leverage AI effectively. If you don't have that foundation,
[00:00:21] Tom Baldwin: My name's Tom Baldwin. This is Overruled By Data, the podcast for law firms looking to start their data journey or accelerate the journey they're already on.
[00:00:28] Tom Baldwin: Brought to you by Entegrata. Today's guests are helping define what modern legal data excellence looks like from building next generation platforms to driving adoption across the business of law. Lisa Mayo Haynes is the Director of Technology Innovation at Ballard Spahr, leading the design and deployment of secure intuitive tools that create measurable impact for lawyers and clients.
[00:00:53] Tom Baldwin: Under her direction, the firm's advanced from a traditional enterprise data warehouse to a modern lakehouse platform, powering analytics, self-service, and AI across the Ballard 360 ecosystem, and a suite of generative AI applications, including Ask Ellis. Ballard, X-Ray and KYC, and then KYC capability.
[00:01:11] Tom Baldwin: Her leadership has earned global recognition, the financial, uh, times innovative lawyers, north American entrepreneur, finalist in 2024, corporate council, women influence and power and law, innovative leadership, honoree legal Intelligence or innovative award finalist and Ilta Trailblazer Award for 2025.
[00:01:30] Tom Baldwin: And the Ballard Spahr silo AI collaboration. Microsoft profiled her team for saving over $2 million in unplanned losses, cutting proposal times down by two hours, and trimming research time by 60%. Bloomberg credited Bloomberg Law credited the program with driving 30% profitability growth. From 2019 to 2022, joining her is Ian Lopez, whose work centers around increasing efficiency and finding the best solutions for lawyers at Ballard.
[00:01:58] Tom Baldwin: This runs the gamut from creating dashboards and surface key insights to managing the generative AI platforms. If the firm has bought and built ensuring that lawyers have access to the correct and most efficient tools for their use cases. Together, Lisa and Ian represent two very complimentary sides of the legal data transformation product and platform innovation.
[00:02:18] Tom Baldwin: On one hand. And data strategy and adoption on the other hand. Whew. Lisa and Ian, welcome to Overruled By Data.
[00:02:27] Lisa Mayo Haynes: Thank you. We're glad to be here.
[00:02:30] Tom Baldwin: So Lisa, let's start with you. Your path from technologist to director of technology and innovation at Ballard Spahr spanned data management, analytics, and now ai.
[00:02:40] Tom Baldwin: What connects those chapters?
[00:02:43] Lisa Mayo Haynes: Well, even though it's a very long journey, decades long, it's, it all is, has been kind of a logical connection. I started my technology career as a developer and developers always have their own ideas about how data should best be organized. At that time, there was no one at the firm that was performing the DBA function.
[00:03:07] Lisa Mayo Haynes: So I started to organize our data in the way I saw that it, it should be, and I later asked for my title to be changed. So from there I progressed to actually leading. Even though I had a DBA Title I, I, uh, led a group of developers in a separate department. And then fast forward to 2018, when we merged with Lindquist and Vennum and I moved into the innovation department.
[00:03:34] Lisa Mayo Haynes: Hmm. And that's where we really had the greatest traction with our innovation efforts around data and analytics. And now in my technology innovation role, I'm at the forefront of some really exciting change.
[00:03:49] Tom Baldwin: I want to circle back to the origin story for you in a second. But Ian, tell us a little bit about how you entered legal data.
[00:03:55] Tom Baldwin: It's not like something, uh, people say, when I grow up, I want to do this for a living. And what first drew you to the power of analytics?
[00:04:03] Ian Lopez: Yeah. No, it's definitely, I sort of stumbled into it, but I was always really interested in. When you drill down and really got into the data, how it affected people and outcomes, right.
[00:04:14] Ian Lopez: It's not just numbers, honestly. Math is actually not my strongest suit, but the analytics piece of it and sort of the trend monitoring that I find really interesting and you can apply it in so many different ways. And so I started at a law firm prior to working at Ballard and then, uh, was hired. Here, and it just really opened my eyes to how much more, um, we could do with data and analytics, dashboards, and really thinking about it from a, from an attorney's perspective and from a business of law perspective.
[00:04:51] Ian Lopez: What numbers and visualizations can we put in front of. The people who need to make a decision quickly, we have time to dig into it. They don't. And so, um, just sort of trying to marry the data with the, with the business of law, uh, was really interesting to me. And of course, as Lisa said, now that AI has come on the scene, it's a whole new world that where, you know, engaging with and sort of trying to work into the, the culture at the firm.
[00:05:21] Tom Baldwin: So Lisa, you said something interesting. You started off as a developer doing DBA work. You kind of fell into it. There was probably this moment, and if you could share the details, that would be always fun. Is there, there's always a moment where someone like has this, uh, epiphany or slash panic attack and they're like, I don't have what I need to do, the thing I'm being asked to do.
[00:05:44] Tom Baldwin: What, what was that moment for you when you were at the firm that you realized the power or the, the lack of having data created? Anxiety.
[00:05:51] Lisa Mayo Haynes: Yeah, I realized that we did not have data that was always accurate, relevant, timely, and complete, and that created bottlenecks for firm management because they had reports that maybe the financial team had to reissue because there was a mistake or something in the system behaved in a way we didn't expect.
[00:06:14] Lisa Mayo Haynes: Um, or maybe there were duplicate data points, for instance. Mm-hmm. And so because there were duplicate data points, they didn't know which one was complete and or which one was correct. And so that really led to problems. If firm management can't trust the data. That you're giving it or that the accounting and and financial team is giving it, then that's a serious problem.
[00:06:36] Lisa Mayo Haynes: And so that's kind of when we, we knew we had to do something, something different.
[00:06:42] Tom Baldwin: And Ian, same question for you. W was there some sort of, sort of like a, a project or a moment of time where you've realized how important it was? And A, a sub B to that is someone who's, who's, you know, self identified as not being comfortable with math, but yet here you are working with data.
[00:06:59] Tom Baldwin: I wonder if that actually helps you a little bit. 'cause a lot of lawyers would say that they also probably are not the most comfortable around numbers. Does that help you in positioning what you're delivering to them in a way that's more bite-sized and, uh, easy to digest?
[00:07:12] Ian Lopez: Yes, definitely. I think it also helped more in the sense of being okay with asking for help.
[00:07:18] Ian Lopez: I think in this, in the data world, right? It's okay to not understand something and so. I always say, you know, there are smarter people in the room than me. And so really bringing different groups at the firm in under one umbrella to do the project. I think the math deficiency, which I'm not terrible at it, it's just not my strongest, but um, yeah, has, uh, has allowed, I think our team in general, which we're really good at, is connecting with different groups and bringing them in.
[00:07:46] Ian Lopez: I think what I realized was, and. I'll give a ton of credit to Lisa and the team. Before I joined, they had built out this EDW, the enterprise data warehouse. And so that was all done, uh, when I came on the scene. And so I realized that we were now able to be predictive in our data and analytics instead of reactive using data to say, this is what we're expecting to happen.
[00:08:19] Ian Lopez: Versus getting it right now and having to react. Um, and that really sort of changed the way that I saw things because it allowed us to help people prepare for good and bad versus having to constantly play catch up. And we, we've extended that. To our clients as well through dashboards and analytics for them.
[00:08:46] Tom Baldwin: Lisa, maybe you could give us a snapshot of Bower's client value and innovation function.
[00:08:51] Lisa Mayo Haynes: The enterprise data warehouse is definitely the foundation for that program, whether we're looking at pricing and alternative fee arrangements, or we're looking at reporting or dashboards that we give to our clients that provide trending.
[00:09:10] Lisa Mayo Haynes: And analysis and things like that. And so that enterprise data warehouse is really the function we were creating that didn't know it was happening, but we were creating our EDW at the same time as foundations was being promoted. So they're very similar tools. The differences. We built ours in-house, but everything that the client value and innovation function creates or uses, leverages that strong data foundation.
[00:09:39] Lisa Mayo Haynes: And we've, we've most recently converted that to a Fabric lakehouse.
[00:09:44] Tom Baldwin: Yeah. And I wanna circle back to that in a little bit because there's very few firms that haven't done some sort of data warehousing initiative, whether it's a small subset of people data, or maybe a more well established EEW that you build.
[00:09:59] Tom Baldwin: And I wanna circle back that in a little bit to kind of double click on your decision to. Move to that platform because for a lot of firms, they're all kind of in this mode. Like, Hey, what we have works and I know how to manage it. I've, I've been a SQL DBA for 20 years and this stuff makes sense to me.
[00:10:15] Tom Baldwin: This fabric stuff is new or Databricks, right? It's all net new stuff. So I'd love to circle back to that 'cause, and we will in a second. Ian, from your standpoint, what's the structure or partnership for cross departmental data? Uh, how's that successful?
[00:10:29] Ian Lopez: So I'll actually say that we're. Very good with integrating with the other teams, with the sort of more analytic data focused team, even with marketing.
[00:10:41] Ian Lopez: So it, it's really just about in kind of communicating, um, we set weekly or monthly sort of check-ins with the different departments. It this question of how do you work cross function functionally, I don't think is as. Complicated as it needs to be. You know, it, it's sort of a simple open line of communication and really helping each other, right?
[00:11:05] Ian Lopez: I know that we can go to the other team to fall that are parallel to us, and they'll be able to assist. The more challenging piece has been how to interface with the lawyers because they, and it's, it's, I don't blame them, right? They're every minute. Counts for them and how to be most efficient in interacting with them, in delivering what we built in getting the requirements, requirements, gathering with them.
[00:11:37] Ian Lopez: That has been a real fun, um, but definitely challenging lesson in engaging with them in a way that that works best for them.
[00:11:47] Tom Baldwin: We're gonna dive into that a bit more because I think one of the missing pieces in most data programs is thinking about it in isolation. That it's owned by it, when in reality it's the business and you all, I love the fact you're so focused on the business of law.
[00:12:01] Tom Baldwin: 'cause I think in most firms it's heavily skewed towards the practice, which makes perfect sense. But it's, it's sort of like the redhead stepchild is the business side, which. Yeah, there's lots of use cases there, right? I But you need collaboration with all the groups between hr, finance, marketing, professional development, km it, everybody.
[00:12:21] Tom Baldwin: Yeah. But before we do do that though, I wanna circle back to Lisa's comment about that shift from your EDW to a a Lakehouse architecture. Lots of, many, many firms we talk to are kind of. Struggling with this internal debate. We've got something that kind of works. Why would we change? What? What capabilities did that unlock for you?
[00:12:41] Lisa Mayo Haynes: Oh my goodness. Loads of capabilities. So even before generative ai, we knew that we wanted to do more with our data. And be able to give our lawyers predictions about what may happen with their book of business. And so our lakehouse really the advantage, the main advantage is number one, it allows us to have structured data and unstructured data in the same repository.
[00:13:08] Lisa Mayo Haynes: We can now take all of that and join it together. So we may have. This unstructured data that's coming from some API or some website that now we can marry with our client data and present in a single dashboard. That's the beauty of what we were doing. And then of course, having it in a lakehouse is also now unlocking artificial intelligence and machine learning capabilities.
[00:13:33] Lisa Mayo Haynes: It's allowing us to do things like client segmentation, which is what we're doing with our know your client. Solution. It allows us to find patterns and trends that humans can't derive on our own. Um, we can also now, um, chat with our data with the advent of generative ai. You know, we can also add additional features to it and we don't have to worry about some of the constraints, previous constraints, like SQL updates and things like that.
[00:14:02] Lisa Mayo Haynes: So it's really. Unlocked a whole new suite of, of opportunities for us as a firm.
[00:14:10] Tom Baldwin: Lisa, what would you say to somebody who is really struggling with this decision? You might have had that own kind of thought bubble, like, boy, do I really want to kind of cra like, my baby is just fine. Are you telling me my baby's ugly?
[00:14:24] Tom Baldwin: Like, no. Like, this is great. I, I built this thing. It's been around for 15 years. I know how to use it. I know how to manipulate it. It's good. It does everything it's supposed to.
[00:14:34] Lisa Mayo Haynes: No, we knew that it had issues. Um, it had di issues with effective dated information. As I mentioned, we had duplicate data fields in different systems.
[00:14:47] Lisa Mayo Haynes: So, you know, for instance, per a person's law school class. As the firm sees it, where should that live? Should it be in hr? Should it live in the accounting system? So it was really trying to mediate, I won't say disputes, but disparities where we saw them, and to be able to really have that single source of the truth and that once we had that, that's what made our data so valuable because people could use the data and know.
[00:15:22] Lisa Mayo Haynes: That it was coming from the source of truth and they could have confidence in that data.
[00:15:28] Tom Baldwin: Ian said, predictive, not reactive. I wrote that down 'cause I thought that's really smart. I think most systems are so backward looking and we're used to that in the market, but that's not what the lawyers need.
[00:15:39] Tom Baldwin: Looking forward. Was there an ever a point in time where you did have to, without naming names and getting in trouble, but was there ever a group that was a little bit harder to work with, where they felt you felt like they were. Maybe not on board with this whole Lakehouse thing, and they like to have their data silo and they didn't want anyone to see it.
[00:15:55] Tom Baldwin: And if so, how'd you overcome that sort of cultural or personal barrier that they, they raised?
[00:16:02] Lisa Mayo Haynes: Well, probably the, the hardest group back then, and it's not the stakeholders, but the individuals within accounting, because we're kind of disrupting everything they know with our systems. You know, they were using a lot of access databases, as I said, that did not calculate effective, dated information properly.
[00:16:24] Lisa Mayo Haynes: These databases were slow, clunky, I mean, just not the actual best use case for access databases. And it was like prying, you know, we had to pry their hands away from, from the D, from access. Um, but once they started to see the value and became. Confident in the data, guess what? We're getting the same, you know, when you run it in, in access and it takes 20 minutes to run this one query, you can come into the financial dashboards and get the same data in 30 seconds, for instance.
[00:16:59] Lisa Mayo Haynes: And so it was, it was really about gaining their trust in the work that we were doing. But yeah, that's just the par for the course. Accounting people have their processes and, and that's that. And, uh. Very often they don't want to change, but we got some very progressive leadership who was all in and, you know, worked with us to, and we worked with her.
[00:17:23] Lisa Mayo Haynes: To convert every single one of those databases into, into either a dashboard or a power BI bookmark that they could save for later on. So we were, we were able to do that with our support.
[00:17:36] Tom Baldwin: And thinking about the early days, was there, were there a one or two, some early kind of outcomes that gave you a little wind in your sails where you could say, Hey look, I know we just did this.
[00:17:47] Tom Baldwin: Here's an example of something that really we couldn't have done before, or used to take 30 minutes and now it's 20 seconds. Were there any early wins that you used to kind of sell it internally?
[00:17:59] Lisa Mayo Haynes: Yeah, definitely, definitely the having timely, relevant, accurate, and complete information. We did have a, a former off the shelf system.
[00:18:10] Lisa Mayo Haynes: Which was kind of self-contained, which meant you couldn't really marry its data with other systems. You couldn't change the processes. One of the failures, of course, was if a lawyer took over a client or matter midyear under this old system, all of that fee revenue went to the new lawyer. And so just imagine the amount of backend work that was involved in in reshuffling.
[00:18:37] Lisa Mayo Haynes: You know, revenue so that it appeared properly for both the new lawyer and the former lawyer so that they both got credit for that. So that was a really big deal. Again, using effective data to point in time data so that if we are looking back in time or looking at current information that we have accuracy as far as who has credit for bringing in this, this fee revenue, and the billings.
[00:19:04] Lisa Mayo Haynes: That was a really big deal once. We got comfortable with that and even, yeah, that was, that was a big deal on the data side, on the generative AI side, I would say like our biggest win has definitely been our, our internally built chatbot. Ask Ellis, because it's just an entree into the world, degenerative ai, and it's making people get very comfortable with the whole idea of prompting.
[00:19:31] Lisa Mayo Haynes: How to interact with these tools and, you know, just trying to fo to continue to foster AI literacy the same way as we focused on data literacy.
[00:19:41] Tom Baldwin: Ian? So Lisa talks a lot about effective dating, which is near and dear to my heart. And certainly anytime you can properly recognize revenue for a partner's book, you're gonna be, be speaking their language.
[00:19:53] Tom Baldwin: How do you ensure those insights translated into decisions for leadership?
[00:19:58] Ian Lopez: I'll use two examples actually. One is internal, uh, that we're working on piloting right now, and it's, uh, to that predictive, uh, note. It's a way of the attorneys being able to see their current book and how much of that have actually come in, and then using that data to predict what the rest of the year is gonna look like.
[00:20:23] Ian Lopez: Obviously there's some times when that's not gonna be all that relevant, right? January's gonna be very difficult. October. At that point, you're pretty aware. But for those months in the middle where you're not sure, we built out a MinMax like variable. So you they can see, okay, this is the worst case and this is the best case, and here's where I'm at in my head.
[00:20:43] Ian Lopez: And track where they're going so that they can say, okay, here's where I have to focus. Um, many. It's just for them, right? It's not a comparison or anything like that, but it allows them to see not only where am I, but what can happen if I do X, Y, Z. Um, and that's that sort of predictive piece, right? We are not just saying, okay, you are at 60% of your goal and it's halfway through the year.
[00:21:11] Ian Lopez: We're saying here, 60% of your goal is halfway through the year, and here's where you can be with these various changes. And then we've done the same thing in our legislative dashboard for clients. It's not financial, but we brought in legislation data so that they can be prepared so that they are, they can see what is coming out, what Congress is passing and adjust accordingly.
[00:21:36] Ian Lopez: Uh, so that they are not in breach, so they don't get sued. Right. We're working to prevent that from happening and that that's just part of what Ballard offers, um, to our clients as a, as a value add. And it's, it's got the same tools that, you know, that show the trend in the analysis and build coming out.
[00:21:54] Ian Lopez: And you can drill down by status of the bill. You can drill down by topics so that you can really get a big look at, look out there, what's happening and how you need to pivot potentially, uh, as a business. Yeah. Awesome.
[00:22:12] Tom Baldwin: Okay. Now we're, we're pivoting into the show where we'd like to talk about lessons learned, kind of opportunities for things to do differently.
[00:22:18] Tom Baldwin: During, during, during the implementation of your lakehouse. Was there an a aspect of it or something you didn't land cred the way you wanted? Were there any assumptions you made or how did you adapt your process either during the deployment or post go live?
[00:22:32] Lisa Mayo Haynes: Well, I'll tell you what ended up happening. We had just started to move our EDW to the cloud.
[00:22:39] Lisa Mayo Haynes: Working with a, a trusted, uh, technology partner and everything was going well. And then about a month or two in, we found out that Microsoft was launching Fabric. We weren't going that route 'cause there was no Fabric. We started reading about it and you know, everything we've read said if you're in the very early stages of converting or you haven't started.
[00:23:07] Lisa Mayo Haynes: Pivot and go with the Fabric route. And that was huge for us to actually make a decision, like, we are going to stop what we're doing and we're gonna go this way. But I think I know now that we made the right decision. So we, we, we would've been fine as is using Azure Synapse and, and all of that. But the way that Fabric is marrying the data engineering and the data science and.
[00:23:36] Lisa Mayo Haynes: Data, architecture, all of that into one tool really is, is huge for us. I mean, it it, it wasn't that we were on the wrong track, but we had to quickly a better, a better
[00:23:48] Tom Baldwin: track emerged for you.
[00:23:50] Lisa Mayo Haynes: Exactly. Yeah. Exactly. And then, and being not afraid to say, okay, we are gonna go this way.
[00:23:56] Tom Baldwin: So in that process, you, you go to Fabric, you're an early, early, early adopter.
[00:24:01] Tom Baldwin: Mm-hmm. Um, whether it's Fabric or snowflake or databricks, if you take sort of the tooling out of it, was there anything along the way that if you were to go start this project again from scratch, that you would say, Hey, we did it this way and it worked out. We got to where we needed to go, but doing this over, I would do this differently.
[00:24:21] Lisa Mayo Haynes: I think the one thing we would do is we built out our semantic. Uh, layers. We created a lot of measures that were very specifically time centered. So a five year revenue number, a three year revenue number. And you know, as the technology evolved both through Power BI and Fabric, you know, we realized we didn't have to do that.
[00:24:46] Lisa Mayo Haynes: We could have just, ugh. You know, had the metrics in there and then we decided, okay, we want to see these metrics for year to date, for going back one year without having to explicitly create create.
[00:25:01] Tom Baldwin: Yeah. And was that sort of a, a carryover from just the old way you had to do things in a Yeah, we really EW Yeah.
[00:25:08] Lisa Mayo Haynes: Yes, yes. But really impactful in terms of speed to delivery, to not have to spell out all of those different fields, so that, that was a really big deal for us.
[00:25:23] Tom Baldwin: You often say adoption hinges on data confidence. Do you have any tactics for creating trust in numbers amongst the partnership?
[00:25:32] Lisa Mayo Haynes: Yeah, unfortunately, you have to just put the work in.
[00:25:35] Lisa Mayo Haynes: We took a very deliberate process when we first rolled out our financial dashboards. We wanted every timekeeper starting with the partners to be data literate and understand what they were seeing and how to get to what they needed. So I would actually sit with them, each one of them individually and find out.
[00:25:56] Lisa Mayo Haynes: First question I would ask is, what data do you need to see? On a daily basis. And then we would go through and I would show them how to get to that answer. And if it wasn't something that was readily apparent, I would create Power BI shortcuts for them so that they could, would know, okay, just go into this menu item and click and quickly find insights.
[00:26:21] Lisa Mayo Haynes: And like I said, it's a lot of work. But when it comes to like compensation, when you're talking about allocations and the, and the drivers of it, partners know their data and so when they were able to answer their questions with confidence and have the data to back it up, that confidence really grew. And you could always tell when we were getting close to allocations, because we started getting either requests.
[00:26:48] Lisa Mayo Haynes: For the same types of reports. But the nice thing is we could show them how to, how to get it themselves. And like I said, they weren't questioning where the numbers came from, so everyone had confidence in, in their data.
[00:27:02] Tom Baldwin: So it sounds like it was really kind of a boots on the ground gorilla warfare, like one, one at a time.
[00:27:08] Tom Baldwin: Partner by partner. Yeah. Sometimes that's the only way to do it, right.
[00:27:13] Lisa Mayo Haynes: Unfortunately. Yes. Yeah. Especially because when you, some partners are basically solo practitioners. Other partners, you know, work with a team, or maybe they're a practice group leader. You add in department chair leaders and then of course firm management.
[00:27:29] Lisa Mayo Haynes: Everyone has a different way of, of looking at these dashboards and, and what's important to them. And so we need to find out again, what do you need on a daily basis? And then we would zero in on that.
[00:27:42] Tom Baldwin: When you start to have this level of access to data, you need to have governance and access controls in place.
[00:27:49] Tom Baldwin: Mm-hmm. So that people are seeing what they're supposed to see, what's proven to be most effective and useful for you in, in governing and, and controlling access inside of the lake house.
[00:28:00] Lisa Mayo Haynes: We have, we actually have a, a great homegrown security mechanism. It's actually a power app that was built by my colleague Steve Magnusson.
[00:28:09] Tom Baldwin: Oh yeah. Steve's amazing. Yeah.
[00:28:11] Lisa Mayo Haynes: He is amazing. Yes, he was a huge part of this. Whole effort. But this security mechanism uses basics about a person, like their job position, their office, their department, their PR, primary practice group to add users automatically to the appropriate Azure security groups. And those security groups drive data access.
[00:28:34] Lisa Mayo Haynes: So we know that if someone is an office managing partner. In the Baltimore office, then we know based on those credentials automatically, who they should be able to see. And then he built the system flexibly enough that, um, we can also enter date-based overrides with expiration dates and things like that.
[00:28:55] Lisa Mayo Haynes: And even an explanation who requested this? Why are we giving this person an override to see data? That they normally would not be able to see. And then of course, we've put a Power BI dashboard on top of that security mechanism. Hmm. So that we could go in, you know, one of the, the huge wins was our chief was asked, who can see?
[00:29:17] Lisa Mayo Haynes: We can see financials at a firm level. She went in all by herself and was able to get that information using the dashboard on top of the security mechanism. And so now we even use it for our custom developed apps as well. Amazing. I mean, we really think about it. Everything is, you know, it's like who's in your little circle?
[00:29:36] Lisa Mayo Haynes: And that's what the security mechanism does, is it, it it automates who's in your circle?
[00:29:43] Tom Baldwin: The observability piece to that is often underlooked unless you've lived that morning or afternoon. When someone comes into your office and asks a very pointed question, and you know why they're asking that, and they're not gonna wanna wait a day.
[00:29:57] Tom Baldwin: Right. Or a week. That's awesome. That is so cool. That's amazing. Okay, switching gears to, we talked about some, Hey, what would you do differently? Let's talk about what you, what y'all are proud of. Ian, what project are you most proud of or something you did that visibly changed the behavior, either of a, a practice, a lawyer, or a client outcome?
[00:30:16] Ian Lopez: So I'm gonna take us a little bit down the, the AI route just a bit. It is, I do think that AI and data are so interconnected, um, that separating them is, is sort of silly. But we have brought on, um, a handful of third party vendors, third party technology partners that have created tools that our attorneys use and.
[00:30:42] Ian Lopez: The process by which I'm really, really happy with the usage, with the feedback, with how excited they've been. They come and say that it saved them time. It's allowed them to prepare for client meetings. We are deploying these tools throughout the firm, so it's not just the attorneys who are using them.
[00:31:04] Ian Lopez: We're allowing LEAs or people in the. Um, you know, the research area, the document, um, repository approval, we're allowing them to use it so that it's, it, it's a full partnership across the firm. But I think the, the best part of that has really been our intake process. Hmm. We do not. Just give out the tool.
[00:31:31] Ian Lopez: So it's really grown organically by word of mouth. And so typically an attorney will come to us and say, what can I use X, Y, Z? And we'll have a 15 minute meeting with them to see what it is they need. Sometimes it is a much more data heavy tool. Something in our Ballard 360 platform that we built before AI was on the sea.
[00:31:53] Ian Lopez: Sometimes it is in fact an AI tool, but we have. Created. Not only did you create relationships with the attorneys, which is awesome, but we are really coming from it, coming at it from a solution driven standpoint. Not just we made this purchase or we built a tool because we thought it would be good. It's what can we do for you?
[00:32:16] Ian Lopez: How can we listen, understand what you need, and then direct you to the right place? Because that's really our job as. Innovators and sort of solution driven business professionals is to get them to the right tool and to understand where they're coming from, not just sign them up for the easiest thing just because we made the purchase.
[00:32:36] Ian Lopez: And that has really informed what we've both purchased externally and built internally, because we will do demos and say, you know what? Our team can build this. We don't need to pay.
[00:32:53] Lisa Mayo Haynes: Mm-hmm.
[00:32:53] Ian Lopez: Yeah. So, and, and we can create this in-house. And so, uh, I think that our, our process is really, really good.
[00:33:02] Tom Baldwin: That's amazing.
[00:33:03] Tom Baldwin: Lisa, same question for you, but I'm gonna narrow down a little bit between Ask Ellis, Ballard, x-Ray, or the Know Your Client initiative, which are you most proud of and why?
[00:33:13] Lisa Mayo Haynes: I, I would say Ask Ellis because that solved the problem of being able to use generative AI against client or firm data without worrying about data leakage.
[00:33:28] Lisa Mayo Haynes: And even if they're not uploading. Uh, client or, or firm data. Everyone writes multiple emails every day. Doesn't matter what your position at the firm is and Ask Ellis just elevates our writing, and I'm especially proud of the full suite of products that we have created. We have draft, Ian has built pre-built prompts so that people don't have to worry about what to say when they're drafting emails or blog posts.
[00:33:58] Lisa Mayo Haynes: Or more. We have the analyze feature, which does let lawyers upload several documents at once and pick whatever output if it's tabular or in a paragraph or bullet point list. These tools are saving our lawyer significant time, doing non-billable work. And the great thing about them is it's a safe entree into the generative AI world.
[00:34:23] Lisa Mayo Haynes: We even tell them to use it for. Personal use, use it if you're planning a trip, a weekend trip, or you know you're going to Europe and you want to see what are the activities that are available. So really, really proud of it. It's, it's a very flexible tool. We're always adding additional assistance, trying to make it better.
[00:34:43] Lisa Mayo Haynes: They can download whatever document or visual or whatever's created that Ask Ellis creates, they can download it and reuse it. So I'm really proud of that because, and, and as far as user adoption, I mean, we, we just hit um, 900 users, so that is over half of the firm. Wow. So that is really good. That's amazing.
[00:35:06] Lisa Mayo Haynes: Yes, it's, we're very proud of the adoption rate with Ask Ellis.
[00:35:11] Ian Lopez: Lisa isn't good at talking about herself, but she had the foresight when AI first sort of exploded to build this on an internal tenant. And so from the jump she saw, we're gonna need to make sure that this stays secure and in our environment.
[00:35:32] Ian Lopez: And because we started from that place with our, we had a consulting firm that we worked with to build this, we were able to be so far ahead of, I think a lot of other firms because we had started from this place of, okay, we are doing security for right, and that's Dave's security. The lawyers won't use it if they don't feel like their data is gonna be safe and secure.
[00:35:55] Ian Lopez: And so that is our. Utmost first top priority, and from the beginning that sort of took center stage and then everything flowed from there and it made it so much easier to have it built to encourage adoption because from the beginning we weren't worried about the security.
[00:36:17] Tom Baldwin: You were mentioning adoption, and firms can define this all kinds of ways.
[00:36:23] Tom Baldwin: How do y'all define adoption?
[00:36:26] Lisa Mayo Haynes: I would say we look at who's using the tool, but then we look at how often they're using the tool. And so we don't just present a metric of, oh, we have this many users on on the tool. We look at who has used it. Maybe they signed in one time and they haven't signed in since, and so we try to follow up with those folks and find out, was it that.
[00:36:52] Lisa Mayo Haynes: Were they just curious? Did it not meet their needs? You know, we try to find out though, we call them irregular users. We try to find out why they did not continue using the tool, and sometimes it's, it's that they need a more needed, a more complicated tool and, and that's okay too. Because we have, like Ian said, you know, we have tools that we've built ourselves, but then of course we have tools that we've invested in also that we've purchased.
[00:37:21] Lisa Mayo Haynes: And so our job really is to find, you know, what is the, the right tool for that lawyer and for his particular use case, his or her particular use case.
[00:37:30] Ian Lopez: And also each tool's different, right? We're not applying the same metric to each one. Hmm. Some tools like Ask Ellis, we'd love them to be using every day.
[00:37:40] Ian Lopez: Other tools that we have, document analysis or upload drafting tools. Maybe they use it once a month, maybe legal research tools or you know, with the caveat of course, that everything has to be checked and verified, but tools like that maybe they're using once a week. So we've worked internally and also worked with the.
[00:38:02] Ian Lopez: Owners, right? The builders of the tool, either internal or external, to figure out, okay, what is our goal here? What are we building this for or buying this for? And then what is the metric based on that? Because to apply one metric across all of them wouldn't make any sense.
[00:38:16] Tom Baldwin: Yeah, right. That's a smart, that's a good, good way to think about it.
[00:38:20] Lisa Mayo Haynes: Mm-hmm. And
[00:38:20] Tom Baldwin: when you think about ROI and usage, those two kind of go hand in hand. So you've, and you've done a, I think we said tremendous job. Of socializing, sort of the bottom line, you know, effectiveness, which is harder to do on certain AI use cases where you don't have a way to quantify it. So how, what metrics do you look for or that you've reported on, and how do you communicate those results, both internally and and sharing externally?
[00:38:49] Lisa Mayo Haynes: When you look at ai, you know, what are the metrics, cost savings, revenue gains, time saved. And client satisfaction. And so that's what we look at. Now, every tool is not gonna meet all four, but for instance, you mentioned Tom, the, the tool we be built with our Ballard X-ray that saves 2 million in, not in unbillable time.
[00:39:14] Lisa Mayo Haynes: And basically it's, it sounds like a lot, but it's really simple. It's two hours per RFP per partner. And if they're, if they're responding, if every partner's responding to. One RFP per quarter, that brings you to that $2 million in savings. So it's really, it sounds small, but it's just an example of how small incremental changes can make a huge difference for your firm.
[00:39:41] Lisa Mayo Haynes: And then, you know, our Zillow, our Zillow product, it's not our product. Of course Zillow is a startup, but we're working with them and that ticks all of the boxes because we're saving money. We're not using tens of of contract attorneys sitting in a room for months. So we're saving money on that front. We are saving time because instead of them sitting in a room for four or five months, we're getting it done in a week or two.
[00:40:09] Lisa Mayo Haynes: With higher accuracy results we're, we were using Zillow to increase our revenue because now we're able to charge a Gen A Gen generative AI fee, and we're saving, and again, I mentioned the time saving. And then of course, the client is saving money. The client is. Our clients are saving hundreds of thousands of dollars using Zillow for complex litigation.
[00:40:32] Lisa Mayo Haynes: So I always love to talk about Zillow because it ticks all of the boxes and it's just a great tool that's been very well received and, and has very concrete ROI that we can associate with it.
[00:40:48] Ian Lopez: And internally, we're working on building out a. A tracker that looks at the usage for all of our tools. Um, obviously as I said before, they'll be different in terms of the goals, the benchmarks that we're looking for, but we'll be able to see this all on one page and we can set our own goals, right?
[00:41:09] Ian Lopez: And then we can show this to our management as. A report to see where we're succeeding, where we need to maybe focus a bit more and maybe we have to say, you know what, this tool's no longer working and that's okay too. We're not tied to any one thing. The only other sort of more nebulous piece I think is we get a lot of qualitative feedback, right?
[00:41:32] Ian Lopez: The quantitative is easy. What we do take into account. An email from a partner saying, this was amazing. I am so happy with this tool. Right? And so engaging with those people who are champions of the various tools and it's helping that to grow organically is also another kind of measurement of ROI that's harder to track, but I would say equally as important.
[00:41:58] Tom Baldwin: So lightning round. Couple more topics here. You mentioned Zillow. What's next on Ballard's AI roadmap and. How are you leveraging the Lakehouse in support of those new initiatives?
[00:42:10] Lisa Mayo Haynes: Great. Great question. So Agen AI is what's next. Some use cases are around client name matching. Matching with our CRM data, but using generative AI to match against client names that are in our lakehouse and do that really, really quickly.
[00:42:30] Lisa Mayo Haynes: Lots of reasons for that, either for leads or government lists where lawyers want us to say is my client, any of my clients on this list where there's literally thousands of entities. So things like that. We're leveraging our lakehouse data paired with ai. To search those clients in just really a couple of minutes.
[00:42:51] Lisa Mayo Haynes: I think Jason has it down to like five minutes now, searching thousands of clients, and so we want to automate the work that first level workers are performing. Not so that they can lose their jobs, but so that they can move on to more valuable, engage and engaging work.
[00:43:09] Tom Baldwin: Where do you see intelligent agents?
[00:43:11] Tom Baldwin: Co-pilots fitting into the workflow of the firm in the next year or two.
[00:43:16] Lisa Mayo Haynes: I think that's going, they're going to be an integral part. The what Microsoft has designed with Microsoft Graph Technology on top of M 365 and co-pilot that is unleashing all kinds of amazing capabilities, including. The integrations that they have with, with multiple systems, ServiceNow, Workday, you name it.
[00:43:37] Lisa Mayo Haynes: And so we look forward to people being able to, um, set up sending emails, waiting for someone to a, to respond more self-service, more quicker feedback, things like that.
[00:43:51] Tom Baldwin: If you could future proof one capability, what would it be?
[00:43:55] Lisa Mayo Haynes: Hmm. I would say it's still all about the data. It's always gonna be about data quality.
[00:44:04] Lisa Mayo Haynes: Um, no matter what you're doing, if it's generative AI or lakehouse data where you, where you're building dashboards on top of it, um, even agents that might be going and grabbing data and reporting that back. If you don't have the trust of your firm with accurate data, then you're really cooked. So. I mean, really good quality data is the foundation of any strategy, an IT strategy, an innovation strategy.
[00:44:32] Lisa Mayo Haynes: You're just not gonna be able to leverage AI effectively if you don't have that foundation. And that's the thing to future proof, in my opinion.
[00:44:42] Tom Baldwin: So, turning that on its head right now, you've got a, let's say you're a, a firm that doesn't have any of this and. They're of the mindset that like, boy, I love this idea of a lakehouse, but our data's not good.
[00:44:57] Tom Baldwin: We need the perfect, I've gotta have a data Strat. I've gotta have all these things in place 'cause I am gonna light at this lakehouse and I'm gonna see bad data and our lawyers won't accept that. Like, what would you say to a firm that's sort of in this mode of like, I need to spend a year or two getting everything perfect and then I could start to look at a platform.
[00:45:17] Lisa Mayo Haynes: I'd say it doesn't have to be perfect. However the financial data should be because right, if you lose them there, then they're not gonna come back. But get your financial data perfect. Get your people data perfect. You know, you wanna know, if you look back five years, what role that person was in, what was their billing rate at that time, and their billable hours Target.
[00:45:42] Lisa Mayo Haynes: Get that right. Once you have that, you can at least start to democratize your data and share it, and you can start to build on. So you don't need a, a lakehouse that has everything under the sun in it, but get the people in the financial data, right? And then you can start to add additional, like I I mentioned unstructured data.
[00:46:07] Lisa Mayo Haynes: You can start to bolt on additional features. Um, as you go, and it should always be evolving, right? I think that's
[00:46:15] Ian Lopez: where the, the difference is, right? The, the data has to be perfect the way that you implement it. That you can move fast and fail fast, and that's okay, right? The way that you're delivering it might not be the best.
[00:46:30] Ian Lopez: And if you change that halfway through. Or you switch it, or the visual looks different, the dashboard looks different, the data in there looks different. That's okay. Right? You, you engage with, with your, with the people who you are giving deliverable to, you take the feedback and you work from there. But if the, if the underlying data itself isn't right, then like Lisa said, you're
[00:46:54] Tom Baldwin: cooked.
[00:46:55] Tom Baldwin: Lisa and Ian, thank you so much for sharing your journey and insights with us today. For our listeners, I hope this conversation reinforces the theme we come back to often on Overruled By Data. Start building your platform and let the strategy grow alongside it. Well, that's it for this episode. If you enjoyed today's conversation, hit that subscribe button and never miss an episode.
[00:47:14] Tom Baldwin: Again, thanks for listening and we'll see you next time on Overruled By Data. Thanks everybody. That's a wrap for this episode of Overruled By Data. If this podcast resonated with you, if you took one or two things away from it, you want to hear more from law firm leaders that have been there and done that hit the fall button.