In this TechVoices interview from CES 2026, MediaMint CEO Rajeev Butani explains how the company has evolved from a traditional services firm into an agentic, AI-powered operations partner for enterprises.
Drawing on MediaMint’s long experience running sales, marketing, media, and data operations for a long list of customers, Butani outlines why agentic AI is fundamentally different from prior technology waves, how human-in-the-loop governance is essential, and why outcomes—not tools—must be the focus.
The conversation also explores MediaMint’s new agentic platform, Mia, the idea of “services as software,” and what enterprise leaders should expect as AI agents become embedded directly into workflows.
Key Themes from the Conversation
- Agentic AI is about outcomes, not tools. Unlike SaaS or cloud, agentic technology is designed to perform work and deliver results, requiring companies to rethink how they adopt and govern AI.
- Human-in-the-loop remains essential. MediaMint views agentic AI as an amplifier of human capability, with accountability and decision-making ultimately remaining with human experts.
- Mia is built on deep operational experience. MediaMint’s agentic platform uses persona-based agents grounded in real-world roles like media planners, campaign managers, and yield analysts.
- “Service as software” defines the future. Enterprises increasingly want partners who take responsibility for outcomes, not vendors who simply deploy tools and walk away.
Important Quotes
Agentic AI Focuses on Work, Not Just Technology
“When we think of agentic, this technology is focusing on work and the end outcome versus just providing a tool and walking away. That’s what makes it very different from prior waves like SaaS or cloud.”
Readiness Matters More Than the Model
“It’s less about the foundation models and more about the readiness of the company to adopt them—its data, its operational maturity, and how fragmented its systems and tools are.”
Agents Amplify Humans, They Don’t Replace Them
“We don’t look at agentic as autonomous. We look at agentic as amplifying what humans are doing, with humans in the loop and at the center, accountable for the outcome.”
Service as Software Is About Accountability
“Customers want organizations that can drive end outcomes for them, not just provide a tool. That’s what we mean by productizing a service using agentic AI.”
