The push toward intelligent automation is accelerating as enterprises realize that the real breakthroughs in artificial intelligence now hinge on how well systems can fuse data, context and autonomous decision-making into a single operational engine.

That shift is redefining what it means to build value with agentic AI, pushing vendors to prove they can handle the messy, unstructured data that fuels autonomous work. Elasticsearch B.V. is leaning into that pressure, framing itself as the backbone for a rising “agent cloud” where context isn’t optional — it’s the whole game, according to Ash Kulkarni (pictured), chief executive officer of Elastic.

“In the world of AI, it’s all about context engineering,” he said. “It’s all about providing context to these large language models so they can actually do their job correctly. At AWS, the announcements that we made this week have all been around that same theme. The first thing that we announced was our integration with AWS AgentCore.”

Kulkarni spoke with John Furrier at AWS re:Invent, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed Elastic’s blueprint for multibillion-dollar scale as enterprises shift toward the agentic era.

Intelligent automation is the new AI differentiator

As much as 80% to 90% of enterprise AI benefits will come from agent systems — not just models, Kulkarni noted. For Elastic, this is a massive opportunity. Its long-standing expertise in unlocking value from unstructured data — documents, logs, messages, records and transcripts — is now mission-critical. In the intelligent automation era, everything comes down to context engineering: giving models the right private data at the right time via retrieval, search and intelligent context pipes.

“At the end of the day, it doesn’t matter if it’s a Nova model, whether it’s a Gemini model, whether it’s an OpenAI model or an open-source model like Llama or Mistral, all of these models are trained on publicly available data sets,” Kulkarni said. “They really have no context about what’s private to your business. The task of context engineering is connecting the dots between all of these models and your private data.”

To reach that agent cloud vision, Elastic has invested in several key partnerships and integrations. Besides the AgentCore tie-in, a new partnership now combines Elastic’s platform with Accenture plc’s services expertise, enabling enterprises to build real-world agentic workflows — now available directly through the AWS Marketplace, according to Kulkarni.

“The second thing that we announced was that we were awarded for AI Competency,” he said. “We’re one of the first ISVs that AWS awarded this competency … just based on all the work that we’ve been doing in the AI space around retrieval augmented generation, around context engineering, around agent development.”

The shifting architecture landscape means that a new cloud paradigm is emerging around data systems, not servers or infrastructure. As competition increases, costs will drop, enterprises will demand flexibility and model choice and the true differentiator becomes the data layer, not the LLM, according to Kulkarni.

“I think the new platform is going to be these data systems … the data platforms,” he said. “I do agree that there’s going to be an emergence of a new type of cloud, but in my mind, it’s going to be a data cloud. The role that we are very determined to lead is the role of being this data cloud for unstructured, messy enterprise data that has the greatest value in it.”

Here’s the complete video interview, part of SiliconANGLE’s and theCUBE’s coverage of AWS re:Invent:

Photo: SiliconANGLE