Videos
Confluent on Real-Time Data, Agentic AI, and Enterprise Intelligence
Insights
- Agentic AI becomes effective only when powered by real-time, contextual, and governed data streams.
- Enterprises are shifting from batch processing to low-latency, source-driven intelligence to improve speed and reduce infrastructure costs.
- The next phase of AI transformation will depend on trusted ecosystems that connect data platforms, governance, and business outcomes at scale.
At MWC 2026, Monish Sharma of Confluent explores how enterprises are rethinking data infrastructure to support real-time AI-driven operations. The conversation highlights how agentic AI depends on timely, contextual, and governed data streams rather than static batch processing. He explains how organizations are shifting intelligence closer to the source, enabling inline processing, lower latency, and more efficient decision-making before data reaches expensive storage environments. The discussion also addresses the growing importance of governance, traceability, and centralized control as enterprises scale AI adoption.
Monish Sharma:
So if you think about AI, AI can only be useful and effective if it's got valid data that it acts on. What Confluent does is, as we connect into the different data sources, you can pick and choose the data that is relevant to your agentic AI outcome and provide that in real-time contextual to that use case for that time to that agentic AI.
They can then make meaningful decisions, relevant decisions, relevant outcomes and feed it back into the source where you want to interact back with the customer.
Enterprises are moving from batch to real-time decisions
Monish Sharma:
So that means role-based access control, lineage, auditability, traceability, who's connecting to what, at what time, what are they accessing? All of that information being provided through a central control dashboard, which is our command center. And then we stream that data from that specific source in terms of the Apache topic to the data to the target location where you want that to be taken. And while we're streaming it, we can, we also have the capability of inline processing of that data.
So you don't have to really wait for that data to be processed at the data lake house there, you know, where you have expensive storage and compute and larger storage to go through.
So we can drive more and more of that processing towards the source, saving cost, saving time and doing it in very low latency.
Frontier telcos optimize cost through data efficiency
Monish Sharma:
We also drive greater efficiency in terms of the amount of data that we can save from going into a data lake house because we do that processing more towards the source. That's what we call the shift left aspect of our business.
And by doing that, we're only pushing data that is relevant into the data lake house, which has, which is expensive in terms of the kind of storage that they have.And then there's also cascading costs that we save because the more storage you have in the data lake house, the more compute power you have to consume.
And then there's also the CSPs, the cloud service providers also charge the egress costs for that data, which we also save. So those cascading costs is what we end up saving for the customer. And so at enterprise scale, at telco scale that is huge.
And so we're really enabling frontier telcos by really operationalizing this, you know, in real time and driving that intelligence at scale.
Partnership turns data into business outcomes
Telecommunication companies and enterprises, they have lots of complex challenges. They need a trusted advisor that can drive that road map for them. Bring the rest will be technology that really fits the environment and also drives the outcomes for those customers in terms of total cost of ownership savings, the productivity gains, the outcomes that actually help the customer experiences.
And that is why I think Confluent and Infosys have a really good story to tell. And there's a symbolistic play there where, you know, Infosys enabled the Confluent data streaming platform technology can drive those outcomes for our biggest enterprise companies and customers.
