Can Confluent Intelligence solve the biggest data challenges in AI?

“I think that to truly capture the power of AI, there's just no way around this. You really have to have to figure out how to get at your data and how to bring it to bear for your problems,” explains Shaun Clowes, Chief Product Officer at Confluent.

As businesses aim for faster AI development and deployment, Confluent Intelligence is expected to provide everything organizations need to launch and scale their AI systems. Unveiled in October 2025, the fully managed stack that is built on the Confluent Cloud, continuously streams and processes historic and real-time data, delivering this context directly into AI applications for more reliable, secure, and highly scalable AI workloads.

In an interview with CRN Asia, Shaun Clowes, Chief Product Officer at Confluent explained that Confluent Intelligence extends a mission that the data streaming platform has been on for a very long time. With enterprises constantly challenged with moving information freely across their estate, Clowes added that AI has only compounded the problem even more.

As such, organizations not only need a wide view of data but also need data from across all of the operational world, which is extremely challenging for most organizations today.

“Data is fragmented and shared across many different operational databases. In our data streaming report in 2025, 70% of APAC businesses cited data sprawl as the main challenge to accelerating AI and ML adoption. So, we know that that's a profound problem. We also know that AI projects are stalled, and they're stuck because they’re using models that were trained on history, or if we want them to deliver a good outcome, we need to deliver them all that fragmented data in real time at high quality. So, we're kind of stuck with either bad answers or good answers with data we can't get at,” Clowes explained.

At the same time, Clowes pointed out that businesses are also under pressure to deliver real time intelligent applications but can't do that because they don't have a platform to do it safely at scale in a comprehensive way.

“So, what does Confluent Intelligence do? Confluent Intelligence extends Confluent Cloud so that it makes those problems go away. You can get access to all the data from where it's been created. You can get in real time in high quality, and then you can apply it to all of your different use cases much, much faster. Confluent Intelligence is kind of a cherry on top of the platform. And it makes that really straightforward of this journey we've been on for over 10 years now,” Clowes said.

For organizations, Clowes pointed out that as Confluent has always been a very open data company, Confluent Intelligence is powered by Flink, which is an open data system. This means it can fit into any complicated, heterogenous environment.

“We connect to everything. In a way we're a bit of a Switzerland for data. We help people move their data from wherever it is to wherever it needs to go. And Confluent Intelligence benefits from that. It takes advantage of all of this moving data and bring it to you,” Clowes added.

Getting AI right

According to Clowes, most customers, be it in the APAC region or globally, are using Kafka and Confluent’s technology to move data inside an enterprise in real time for all sort of different business use cases.

“Customers are asking us if they are already using Confluent to move all of this data, to understand it, to capture it. Can we also help them also literally just put it directly into LLMs, into agents, or into their different intelligent applications? So Confluent Intelligence allows them to take advantage of the data that they already have, the capabilities they already have, and immediately just leverage those. So, customers have been asking us directly about this. And Confluent Intelligence means that they can go from kind of zero to a functioning AI agent in minutes,” he said.

Clowes explained that the agent will still be a good one and not a naïve agent that has no idea about the particular company. Put simply, companies can build very good agents quickly because they have access to all the high-quality data.

“We work with all the major model providers. We work with all the major sources of data, all the major sinks of data where we're trying to bring, trying to unify this estate so that customers can get the business outcomes they're looking for. If you look at a lot of the pilots that we work with and a lot of the kind of evolution we see in the industry, most of these AI pilots, they're really struggling to show any kind of ROI,” Clowes explained.

Clowes pointed out that many of these AI pilots also have zero ROI. However, this is because the pilots are not able to go into production as the demos are built without access to the critical data that the agent would need. As such, the demo is almost destined to fail and organizations are trying to figure out ways to avoid this problem.

“I think that to truly capture the power of AI, there's just no way around this. You really have to figure out how to get your data and how to bring it to bear for your problems,” Clowes said.

Opportunities for partners

In June 2025, Confluent announced plans to invest in its partner ecosystem as well. Since then, the vendor continues to enable its partners, with some being excited about the ongoing projects that Confluent is working on.

“There's a lot of opportunities to build better agents, provide better data to those agents and able to enable those agents to take better action in the world. We've also been working a lot on our technologies like Tableflow and Context Engine. Tableflow allows information to land in all of your favourite analytical databases with zero effort. It can also produce really valuable, deep analytical data into your data warehouse with no effort, no additional work at all. So, we're excited to see that expand across more partners who can deliver more value on top of that data,” Clowes said.

For the Context Engine, Clowes mentioned that the platform is used to surface real time information, enabling queries against real time information that Kafka is already moving. This enables the agents to be smart and take advantage of all that real time data over time.

“We want to make Context Engine available to all sorts of applications. That will enable partners to build even smarter applications that can react to things happening in the real world in really sophisticated ways,” he added.

Clowes also highlighted that Confluent continues to work with a ton of different partners all the way from systems integrators to help them build next generation data driven and data intensive applications all the way to ISV partners and OEM partners that embed Confluent’s technology to build smart applications and enable their customers to build smart applications as well.

“Our US$200 million announcement kind of signifies our engagement with the partner community as a company. Like I said earlier, we are the Switzerland for data. So we want data to be an open system, whereby basically everybody is able to build and create more value on top of that data,” he concluded.