Exclusive: Cloudera's Abhas Ricky on GenAI adoption accelerating fast

Generative AI is no longer a future ambition for global enterprises - it's a reality now delivering measurable gains, according to Cloudera's Chief Strategy Officer, Abhas Ricky.

Speaking exclusively at EVOLVE25 in New York, Ricky responded to a recent MIT study that claimed 95% of GenAI implementations fail due to poor alignment with existing workflows.

While he acknowledged the challenges, Ricky said the findings overlook the real-world successes already unfolding at scale.

"Every study has a level of context," he said. "Everything has to be viewed in terms of what the workflows were, who responded to it, what was the use case that they had to get through."

Ricky painted a much more optimistic picture of GenAI's current enterprise adoption curve.

"When we speak to enterprise customers, they're already starting to see productivity improvement," he said.

At EVOLVE25, which brings together customers, partners and technologists to discuss AI and data innovation, Ricky shared examples of what Cloudera clients are achieving. One case involved a major global oil and gas company, a long-time Cloudera customer, which saved $1 billion in a single year by applying GenAI to a specific business use case.

"I can't name them because of their reasons," Ricky said, "but that just helps you [see] the scale of improvement they're going through."

He also cited another customer using GenAI for research acceleration. That organisation now expects to significantly shorten its time to market - a critical gain in fast-moving industries.

"They believe they should be able to improve the time to market everything," he said. "You are able to do that. It's brilliant to get through with that."

Ricky emphasised that Cloudera is playing a dual role in this shift - not only as a technology partner, but also as a strategic one.

"We are helping customers become a strategic partner of choice for us, but also vice versa," he said.

While acknowledging that many projects still face difficulties, Ricky drew parallels to the early days of cloud computing and SaaS.

"Even a lot of cloud projects failed before," he noted. "Even projects in the SaaS world opened tickets over there."

For Ricky, the core issue isn't that GenAI doesn't work - it's that the final stretch of integration, the so-called "last mile," remains the hardest part.

"If I had to pick one specific area, I would say the last mile for GenAI is the hardest thing to do," he explained. "And that is largely because of the fact that, to the point, there are already a lot of legacy tools."

Legacy infrastructure, often spanning decades, can be a serious barrier to deploying advanced AI workflows. Ricky offered the example of legacy code migration, such as moving from COBOL to Python or even to CUDA libraries optimised for GPU use.

"Somebody has to go and do that," he said. "But you also have to define the workforce. You also have to define the architectural construct."

Another common misunderstanding, he said, is the cost model around GenAI - especially when focused solely on the price of tokens or outputs.

"One of the slides I'm going to present on Thursday will show that when we see these economics for tokens - X cents or X dollars - that's just part of the iceberg," Ricky said. "Eighty-five percent of the iceberg is sitting below that. That's the operational overheads that nobody talks about."

Yet despite these friction points, Ricky said Cloudera is already seeing a growing number of enterprises moving from pilot projects to production-level deployments.

"In general, I'm very bullish on GenAI," he said.

Cloudera, known for its unified platform for data engineering, machine learning and analytics, has focused heavily on making AI accessible where the data lives - whether in the cloud, on-premises or hybrid environments. At EVOLVE25, this theme resonated across sessions, with a sharp emphasis on bringing AI to data, not the other way around.

The event featured customer showcases, hands-on demos and 1:1 meetings with technical experts - part of Cloudera's effort to shift AI adoption from theory to practice.

Reflecting on the past year, Ricky said the company now has clearer alignment between vision and execution.

"When I think about where we were last year and where we are now - there's a vision," he said. "We know where we're going."

For Ricky, it's not just about optimism - it's about evidence. He believes that the companies who invest now in aligning GenAI with their existing systems and operational realities will see transformational returns.

"The majority of the customers will get to GenAI success very soon," he said.