When a live stream fails, viewers don’t blame the network; they blame the brand. Startup delays, bitrate drops and outages translate directly into churn and lost revenue. As events push into 4K and HDR, media companies are confronting an uncomfortable truth: traditional monitoring and operations don’t scale.
Microsoft’s answer is to treat the delivery layer as a continuous source of intelligence, combining cloud infrastructure, unified telemetry and AI to make live workflows more resilient and adaptive. Central to that vision is swXtch.io, a cloud-native live media routing platform built on Azure that captures and acts on telemetry from the precise layer where stream quality is determined.
Simon Crownshaw, worldwide lead for media and entertainment at Microsoft, sees the partnership as a model for AI applied to delivery. “The importance of being able to apply AI in the right way to a media workflow and leveraging swXtch.io’s capabilities to understand infrastructure and data at a foundationally different level is really critical,” he said. “It’s a symbol of what great partnership looks like.”
Owning the Delivery Layer
swXtch.io was incubated inside a New York-based equities exchange and built to solve demanding networking problems in financial services. That engineering discipline, with its low latency, deterministic performance and zero tolerance for failure, transferred directly to live media.
“We’ve built a unique networking offering that establishes us as the only method to move certain kinds of broadcast workflows through the different locations they need to move—including the clouds,” said Geeter Kyrazis, co-founder and head of business strategy.
Most monitoring tools focus on application performance or CDN metrics, important but downstream of where stream quality is shaped. swXtch.io operates at the signal and transport level, generating real-time telemetry from live media flows and extending visibility into the delivery infrastructure itself.
“You might need to take a live sports match and transport it to eight different inference locations across clouds and edge devices, each expecting a different format,” said Kyrazis. “At the core, those are networking problems.”
Turning Telemetry into Intelligence
swXtch.io’s latest innovation, the swXtch AI Router, is a managed service on Azure that routes live streams to AI inference endpoints and feeds the resulting insights back into production workflows. At NAB 2025, swXtch.io and Microsoft demonstrated the full pipeline: the AI Router handling live routing and telemetry capture, Azure AI inference services performing real-time video analysis and Microsoft Fabric, with Fabric Real-Time Intelligence, serving as the “corporate brain” by connecting brand-safety decisions to the contractual data that governs them.
“Think of it as RAG for live inference,” said Kyrazis. “We used Microsoft Fabric to hold a corpus of corporate data, including contractual terms and brand guidelines, and used that as a modifier to the inference on live media streams. For a brand-safety workflow, you’re not just analyzing frames; you’re checking them against what your advertisers agreed to in their contracts.”
Crownshaw framed the stakes plainly. “As you move into higher resolutions and more devices, you need to know what your content looks like at the other end. Is it color correct? Is it buffering incorrectly? Those things directly impact subscriber retention, and you can’t rely on humans alone to keep up.”
AI-Driven Operations at Scale
Large media and satellite providers run thousands of streams simultaneously while service-level objectives, or SLOs, demand four or five nines of reliability. No human team can watch everything. “With AI, you can have all streams monitored, detecting issues and reporting by exception,” said Kyrazis, describing a satellite provider managing black screens, color bars and incorrect language feeds across thousands of concurrent channels.
By feeding swXtch.io’s telemetry into Azure’s AI stack, operations teams can automate triage and root-cause analysis, trigger agentic workflows for remediation and focus human attention where it matters most. “You don’t want operators waiting to hit a stop button,” said Crownshaw. “AI-driven operations let them move to higher-value work.”
A Blueprint for the Intelligent Delivery Stack
Microsoft and swXtch.io are making the case that the delivery layer is as strategically important as any other part of the AI-powered media stack. By unifying routing telemetry, Azure AI services and enterprise knowledge bases through Microsoft Fabric, the partnership gives media organizations a practical path to modernize operations, extend observability to where it matters and apply AI to network resilience and compliance, not just content.
“Everyone is interested in how AI is going to affect media,” said Kyrazis. “What’s missing is an easy way to use it without building your own pipeline. Our goal with Microsoft is to make it easy to get streams to the right inference locations, with the latency and reliability broadcasters need.”
As the swXtch AI Router moves toward general availability and deepens integration with Microsoft Fabric, Azure AI and Copilot, the collaboration is positioning Azure as the end-to-end platform for live media. The delivery layer has always been where streaming brands are won or lost. Now it’s also where they’re built.
About swXtch.io
swXtch.io removes the complexity of networking so media can flow seamlessly into AI systems, from anywhere, on demand. The platform enables customers to route media to any AI inference endpoint across any cloud or specialized providers.
Using natural language, the AI Router translates customer intent into production-ready workflows, automatically designing and deploying the required infrastructure, services and routing with full cost transparency.
Operating over a global, multi-provider network, swXtch.io dynamically optimizes media transport in real time, delivering high-performance, cost-efficient connectivity for live AI workflows, SRT distribution and satellite replacement.
About the Collaboration
swXtch.io is a cloud-native live media routing platform built on Microsoft Azure, delivering broadcast-grade video transport and AI-powered workflows in real time. Together, Microsoft and swXtch.io help media organizations close the observability gap, automate operations and build delivery platforms that learn from every stream.
To learn more about Microsoft’s Media & Entertainment platform capabilities, see microsoft.com/media-entertainment.