Published on: August 25, 2026

Precise and infrastructure-first, Sashi Binani starts from a plain conviction: models are rarely what holds enterprise AI back, the data underneath them is. As Chief Information Officer at LiveRamp and Managing Director, LiveRamp India, Sashi runs a global technology function and a country business at once, and treats both as questions of trust before they are questions of tooling. 

In this exclusive interview, he makes the case for treating data collaboration as an operating model rather than a purchase, sets out what clean rooms really demand of enterprise data, and describes what leading LiveRamp India teaches that the global role cannot.

You’ve held the CIO seat at Dropbox and now at LiveRamp. What changes about the job when the company’s product is data itself? 

I’d frame it a little differently, at LiveRamp, we’re a data collaboration company. What we provide are the capabilities that help brands, publishers, and platforms connect and use data with the governance, permissioning, and interoperability required to drive measurable outcomes. 

That changes the CIO role because the technology becomes even more tightly linked to trust. You’re not only focused on running internal systems efficiently; you’re also responsible for ensuring the company operates with the same discipline that customers expect from the platform itself.

Most enterprises talk about data collaboration as something they buy. What does it look like when it’s how you actually operate? 

When data collaboration is how you operate, it stops being an isolated tool and becomes part of the company’s operating model. You are not standing up one-off projects every time a team wants to work with a partner, measure performance or activate on an insight. The governance, connectivity and repeatable workflows are already in place, so it happens consistently across the business and between clients and partners. 

In practice, that means teams can move faster because agreed-upon rules and use cases are built in from the start, allowing Marketing, Legal, and partner teams to move faster and with confidence that each party’s privacy controls are enforced at every step. It also means you can scale beyond a handful of top clients to enable data collaboration across a broad set of partners, use cases, and markets. 

“The companies that pull ahead are the ones that treat collaboration as a business strategy: trusted, repeatable, and embedded in how decisions get made.”

Clean rooms and identity resolution get discussed as marketing tools. What do they actually demand of a company’s data infrastructure?

“Identity resolution is what makes enterprise data usable across systems.” 

It brings together signals from different sources to create a more accurate, comprehensive, and durable view of the consumer, so data can be connected consistently across clouds, channels, and partners. For that to work, companies need high-quality inputs, consistent identifiers, clear data models, and strong governance around consent, permissions, and data use. 

Clean rooms build on that foundation. They provide a controlled environment where a company can collaborate its data with a partner’s data to generate insights, measurement, and activation opportunities that would be difficult to unlock on either side alone. That requires infrastructure that supports interoperability, secure data handling, access controls, and the ability to connect data without exposing sensitive information or losing trust. 

So while both are often discussed through a marketing lens, they are really a test of enterprise data maturity. If a company’s data cannot be connected with confidence or lacks common governance standards, identity resolution will be incomplete and clean room collaboration will be hard to scale.

Where do enterprise AI programmes most often stall — the technology, the data, or the organisation? 

In most cases, technology advances faster than the enterprise is prepared to absorb. 

“AI programmes usually stall less because of the models themselves, and more because the underlying data and organisational readiness are not yet in place.”

On the data side, many companies are still working through fragmented systems and inconsistent permissioning. On the organisational side, they often have not aligned teams, processes, and accountability around how AI should be deployed responsibly and at scale. 

At the end of the day, AI is only as good as the data beneath it. If the data is incomplete or disconnected, even the most advanced models will struggle to deliver meaningful business outcomes. The companies that move fastest will be the ones that pair strong models with strong data foundations and the operating discipline to put both to work.

You hold a global function and a country business simultaneously. What does running LiveRamp India teach you that the global role doesn’t? 

My global CIO role is primarily about enabling LiveRamp internally by making sure our technology and data infrastructure help LiveRampers do their best work. That role already gives me a broad view across the company, especially at a moment when AI is reshaping how work gets done and raising the bar for what internal technology needs to deliver. 

Running LiveRamp India adds a different dimension. It gives me a much more direct view into how we build talent, scale leaders, and create culture and momentum in a fast-growing environment. 

“Leading India feels like leading a startup inside the company. There is no fixed playbook, so you have to innovate, iterate quickly and build with confidence in the strength of the team around you.” 

What it has reinforced for me is that strong local capability can create real global impact. Our Hyderabad teams go beyond simply supporting the business. They are helping shape initiatives that reach customers and teams around the world. That perspective is hard to fully appreciate from a global functional role alone and running the India business makes it tangible in a much more immediate way.