Evalueserve's Gururaj Bhat on leveraging AI, domain expertise to drive value

Evalueserve named Gururaj Bhat executive vice president to lead its data and AI business. Bhat was previously Data & AI Go-to-Market Solution Leader at Google Cloud and before that was a general manager for Wipro's data, analytics and AI business.

Constellation Insights caught up with Bhat to talk about the new role, components of a winning AI project and why domain expertise and change management matters. Here's a look at the big themes.

Domain expertise as the AI differentiator. Bhat said AI projects to date have been led by horizontal models and cloud hyperscalers that started with AI and then had to add domain expertise often via forward deployed engineers. "What fascinated me about Evalueserve was the deep domain expertise and how we can pivot to be an domain-led AI services company with domain capabilities," said Bhat.

Evalueserve is operating deep workflows within large enterprises. Evalueserve's customer base spans multiple industries and financial services and banking are 50% of the mix with manufacturing and life sciences sizable verticals.

Bhat said domain expertise is key to transforming workflows with AI. "Since we are living in the day-to-day workflows, the ability to enable AI and drive the adoption within the enterprise is where we can bring significant value to customers," said Bhat. "The lack of domain expertise has been the big showstopper in many enterprises."

That domain knowledge is also critical to managing AI costs. If you know the workflows, processes and sub-processes, you can better select models. "Not every workflow and sub-process within the workflow requires frontier model capability," said Bhat. "If you know the workflows you can be well designed from the get-go in terms of designing them to use the right model for the right purpose."

Defining domain-led AI led services. Evalueserve's vision of domain-led AI services combines domain experts, AI engineers and specialized change management. Bhat said:

"In order to enable AI within the enterprise and drive the transformation, you require a combination of the domain expertise, workflow knowledge and the technology expertise to make it real, supported by the change management capability.

When we talk about domain-led AI services, we are saying we are not going to learn the domain on your dime."

Turning automation into transformation. Enterprise AI projects typically start with automation, but need to transition to transformation. "The starting point is the automation because that is where the enterprises will get started, but what I think they will start realizing is now that I have incorporated AI into the workflow, I can re-engineer the process too," said Bhat.

Change management matters. Bhat said Evalueserve's knowledge of specific domains and workflows plays a big role in change management. "It's not a generic change management because understanding the workflow is going to be a bit different for people sitting on the edge of the process," said Bhat.

He noted that Evalueserve has to talk to fund managers and specific roles within the enterprise. "You need to speak their language. Driving the change in a specific workflow is very different from talking general change management."

Developing IP. Bhat said Evalueserve is focusing on areas where it has domain expertise and building AI-first solutions. "We have identified around five to six solutions which are IP-based solutions, which will be AI-first. That's what we're working on," said Bhat, who said use cases would range from indexing to investment banking to credit underwriting. "It will be a licensing model for the starting platform and ensure they use the right models for the right purpose."

What makes AI projects successful? Bhat said successful AI projects have three ingredients:

• Data. "Data is the center of the universe," said Bhat.
• Deep workflow, engineering and change management.
• An enterprise-wide view of AI over isolated use cases. "Many enterprises are jumping directly into use cases, but they often have to change their approach when they move toward broader transformation," said Bhat.