Dye & Durham may be best known as one of Canada’s most established legal technology companies, but behind the scenes, the organization is undergoing a significant evolution. With a global footprint and a growing portfolio of mission-critical software for legal and business professionals, the company is now betting big on analytics to shape its next chapter.

At the center of this transformation is Murat Beshtoev, Director of Global Analytics at Dye & Durham, who joined the company with a mandate to build its analytics function from scratch — not just as a support layer, but as a strategic lever for growth. In this Q&A with Jack Berkowitz, Chief Data Officer at Securiti, Beshtoev shares how he’s navigating rapid technology shifts, from GenAI to agentic AI, while aligning analytics with shareholder expectations and operational demands.

From establishing quick wins that matter to leveraging large language models for warehouse design and delivery in under an hour, Beshtoev unpacks what it truly means to modernize a legacy company with data — and do it fast. This conversation reveals not only the evolving role of AI in analytics delivery but also how cultural, technological, and business integration are key to long-term value creation.

Edited Excerpts

Q: To kick things off — with all the buzz around GenAI and Agentic AI, are these technologies helping solve the same old business problems in smarter ways, or are they opening the door to entirely new challenges and opportunities?

A typical set of issues and challenges remains largely the same. However, the way we deliver solutions has changed significantly. Expectations are different now. With AI, we’re operating on much shorter timelines and that’s becoming the norm across multiple industries. There’s a general expectation that we can leverage this new tech to deliver smarter, faster, and better solutions, all while significantly reducing FTE loads.

It’s also becoming harder to justify adding more talent. So, the economic and business environment is pushing us to work smarter and discover new tools.

And by the way, there are a lot of innovative startups and brilliant people out there constantly pushing the envelope. They’re delivering tools that can dramatically improve processes. We’re not just talking about incremental gains of one or two percent, some of these tools can make processes 10 times faster or five times more effective.

Q: When joining a new organization, how do you assess what’s already in place—whether it’s established processes, culture, or technology—and then effectively integrate your own ideas and improvements to create alignment and bring the team along with you?

I built analytics sort of from scratch for my current organization. We had a complete changeover — new CEO, new board of directors — and with that came new mandates. The organization had never really focused on data and analytics before, largely because it had historically been extremely successful. However, the shareholders believe the company has the potential to be even more successful, and that’s the reason and the drive behind all this change.

Q: As you’re rebuilding, how do you show early wins to keep your CEO and other stakeholders engaged and supportive, especially when you’re still laying the foundation?

I think there are already enough signals out there, you just have to tune in and start listening. Pay attention to the conversations happening around you: what people are talking about, what they’re concerned with, and what problems seem to be emerging. That alone gives you a pretty good sense of where to focus.

It can feel overwhelming in the beginning because there’s so much you could do. So you need to start with some light, informal inquiries. Talk to people, get a sense of what today’s most pressing themes are, and figure out what you can realistically deliver right now.

You might also already have some scenarios or solutions developed from earlier work, leverage those. Bring in what’s relevant and helpful to your client base right away.

And in the early stages, I’d say the focus shouldn’t be on quarterly or even monthly wins, it needs to be weekly. If you cannot deliver that, it’s not going to be very useful.

Q: You often don’t end up where you initially thought you were going, right? As you start delivering week by week, do you find that unexpected areas turn out to be the real levers for driving impact?

I can give you an example, we’re trying to pull different levers at the same time. We decided to add some structure to our data warehouse. Normally, that would mean pulling in engineers, having discussions, and trying to understand everything. But instead, we just got together for 45 minutes. We used large language models, typed in everything we discussed as we went, and then asked the LLM to generate an initial design of the changes we had in mind for the data warehouse.

Once we had that summary, we fed it back into the LLM and it generated the SQL code to actually build the structure. All of that, start to finish, in 45 minutes. So things are definitely improving and moving much faster.

LLMs also prompt you with follow-up questions, things you might not have considered, or gaps in the data conversation that aren’t fully addressed. It’s like having this “super brain” in the room with your small team, helping you iterate fast and come up with solutions in minutes.

CDO Magazine appreciates Murat Beshtoev for sharing his insights with our global community.