Videos
Confluent on Real-Time Data Streaming for Autonomous Telecom Networks
Insights
- Real-time data streaming is becoming essential as telecom operators move beyond batch processing toward AI-driven decision-making.
- Autonomous networks rely on continuous access to fresh data to optimize performance, resolve issues, and improve customer experience.
- AI adoption requires strong data foundations, governance, and integration between operational systems and intelligence platforms.
At MWC 2026, Matias Cascallares, Sudarshan Pathalam, and Mario Yague Yuste from Confluent discuss why real-time data is becoming the foundation of the AI-driven telecom industry. They explain that traditional batch-based approaches can no longer support the speed and responsiveness required for modern customer experiences, network operations, and AI applications. The conversation explores how data streaming platforms enable telecom operators to process information continuously, allowing networks to make autonomous decisions, adapt to changing conditions, and respond to issues in seconds rather than hours. The speakers also highlight the importance of building the right data foundations before scaling AI initiatives, noting that organizations need governed, accessible, and real-time data to unlock meaningful business value. Together with Infosys' industry expertise and AI capabilities, they see real-time data streaming as a critical enabler of autonomous networks, intelligent operations, and future telecom growth.
Matias Cascallares:
Historically speaking, the telco industry has been very batch oriented. And for the requirements that we have nowadays, the batch approach is not enough. You cannot wait to act for the batch to be completed. You need to process the data as soon as the data arrives. This is pretty much what data streaming is about. Data streaming is an unbounded flow of data that you can continuously process and act on it. Think about your mobile phone, when something happens, you receive a notification. Imagine if instead of receiving the notification immediately you need to wait for the next batch execution. Maybe thirty minutes. It doesn't work. You need to get the notification as soon as possible.
Sudarshan Pathalam:
Traditionally you have two domains in an enterprise. If you are a business, you had an operational domain, and you had an analytical domain. The operational domain happens continuously, people buy stuff, people return stuff, people choose, everything happens continuously. In the analytical space, you deal with it in batch. So whatever has happened, you make sense of it twenty-four hours ago. You're clearly lacking that real-time benefit that you can get from a data streaming platform such as Confluent.
Mario Yague Yuste:
Today the telco services are forced to have this kind of real-time decisions because customers are waiting for a very good experience from their side and they have to react on time, and this is money. Telco businesses are investing a lot of money in their infrastructure and they need to monetize it. And in which way they can do it? Having good effectiveness of data treatment, and with this data streaming platform, is the way to solve it.
Mario Yague Yuste:
You can reconfigure the network in an autonomous way in just a matter of seconds or even milliseconds. Just imagine you are trying to have a phone call and you are having some problems — the network will decide by its own, making some changes in the configuration based on the fresh data you have in the platform in real time. If not, you will need to wait for these kinds of windows of time to make decisions. And today you can have the data, process it and make decisions in just a couple of seconds.
Sudarshan Pathalam:
Our customers struggle to adopt technology. Everybody today is at a spot where they have to transform their businesses. If they have to adopt AI, you can't just 'eat' AI, there's some foundational transformation that has to happen in businesses for you to be AI-enabled. I think Infosys is in a very good spot to actually provide those services. They're sitting at a good intersection of technical know-how and how to implement this stuff, and they have a great product with Confluent where it can actually enable the customers and their use cases.
Matias Cascallares:
Confluent is a very, let's say, vertical-agnostic platform. We are widely used in financial services, retail, telco as well. And I think Infosys complements us in the sense that they bring this domain expertise in the telco space and they also complement with specific domain products like Infosys Topaz and others. I think that's a good mix and match.
Mario Yague Yuste:
The telco market or the telco services are very heterogeneous. We are offering the capacity to put all the information in a centralized way, governed, exposed to any kind of external consumer. And with Infosys we have the perfect match, because they have the fresh data in real time and they have the platform and the AI agents that can consume this information and make the telecom market better.
