For decades, business analysts have lived in spreadsheets reconciling data, repeating manual tasks, and fighting for time to focus on real insight. Savant Labs was founded to break this cycle. Positioned as an agentic AI automation platform, the company helps analysts automate repetitive workflows so they can focus on higher-value work.

“We help them automate their work so they can spend time doing bigger and better things,” says Chitrang Shah, Co-Founder and CEO of Savant Labs, kicking off the first part of his interview with Robert Lutton, VP of Sales and Marketing at Sandhill Consultants.

But Shah’s perspective stretches far beyond tooling. In his view, today’s analytics leaders stand at the center of a pendulum swing — one that has major implications for agility, governance, and how enterprises unlock value from data.

From Excel to AI: A pendulum in motion

Shah describes the history of analytics delivery as a back-and-forth between empowerment and control.

“In the early 2010s, everybody was doing analytics in Excel,” he recalls. The rise of self-service tools promised freedom for business users, but it quickly introduced chaos. “It created a whole host of governance challenges for data leaders,” Shah notes. Data lived on desktops, accuracy was in doubt, and visibility all but disappeared.

Cloud data warehouses marked the next shift. Vendors like Snowflake and Databricks urged enterprises to centralize analytics into a governed environment with clean, standardized datasets. Many organizations embraced this vision, some going so far as to lock down self-service completely. But this model had its own limits.

“The demand for analytics kept growing, but the people who can build and automate analytics, namely the data engineers, were far and few in between. Organizations could not scale the centralized approach,” Shah explains.

Now, agentic AI is tilting the pendulum again, toward what Shah calls a new generation of self-service: “Data leaders are looking for a solution that enables self-service but with centralized governance and control. The pendulum is trying to come back to somewhere in between, where you can find the best of both worlds.”

The case for Centers of Excellence

Finding that equilibrium is not a matter of technology alone. Shah emphasizes the triad of people, process, and tools.

“For smaller organizations, a centralized analytics team will make sense, but as complexity grows, self-service becomes inevitable,” he says. “That’s where a Center of Excellence can provide the discipline — defining processes while ensuring the tools enforce governance but also empower.”

In this model, technology shifts from being the constraint to being the catalyst for transformation. “Anytime you’re going from one extreme to another, the biggest problem is change management,” Shah points out. “Having the right technology allows you to instrument change management in a thoughtful, systematic way.”

He encourages CDOs to treat tooling as the first step in scaling responsibly: pilot self-service with a few trusted users, build governance practices around their successes, and then expand deliberately.

Rethinking self-service in the AI era

If the first wave of self-service analytics was about drag-and-drop workflows, Shah argues the next must be about agentic AI. He outlines three qualities that separate effective platforms from the rest:

1. Ease of use: “If an average analyst who lives in Excel can’t use the tool, it’s probably not the right fit.”
2. Democratization of AI: “The next generation tools have to have agentic AI built in. But the tool also has to integrate with your AI to govern and make sure your data privacy is in place.”
3. Scalability: “The cost can’t grow linearly with usage, otherwise it becomes prohibitive to scale self-service.”

In Shah’s view, these are not nice-to-haves but prerequisites for analytics leaders who want to make data-driven decision-making both agile and trusted.

Shah’s pendulum metaphor is more than history; it’s a roadmap. The future of analytics won’t belong to organizations that swing hard to one extreme but to those that strike a balance: empowering business users while protecting data quality, leveraging AI while ensuring trust, scaling operations without losing agility.

As Shah concludes, “Data leaders are trying to find the best of both worlds.”

CDO Magazine appreciates Chitrang Shah for sharing his insights with our global community.