Customer Story
How HPE InfoSight ingests data from 20 billion+ sensors to predict and prevent infrastructure failures.
HPE built InfoSight on Akka to process trillions of telemetry metrics daily from 20 billion+ sensors — petabytes of data feeding predictive AI that resolves infrastructure issues before customers are affected.
Akka ROI Scorecard
issue resolution — problems detected and fixed before customers notice them.
of telemetry data processed and analyzed continuously with efficient resource use.
sensors feeding trillions of metrics daily into the InfoSight AI platform.
Business outcome: the majority of support cases predicted and resolved automatically.
HPE’s InfoSight product collects telemetry from over 20 billion sensors across its global installed base of storage, compute, and networking infrastructure. The volume is staggering — trillions of data points daily, measured in petabytes.
“We have something like 1.2 million systems calling home with over 20 billion sensors,” says Jeff Dutton, VP of Engineering. “The amount of data coming in is just enormous.” The system needed to ingest, process, and analyze this data in real time, correlate patterns across the global fleet, and trigger automated remediation — all with the reliability that enterprise customers demand from their infrastructure provider.
Akka’s streaming capabilities gave HPE the ability to build a globally distributed telemetry pipeline that scales with the installed base. “We really couldn’t find another technology that would be able to scale the way we needed it to. We looked at a lot of different technologies,” says Dutton.
Each sensor stream is processed independently, with Akka handling backpressure, failure recovery, and state management across the distributed pipeline. The result is a system that grows naturally as HPE ships more hardware — without requiring architectural redesigns.
Speed to production. Predictive analytics that identify issues before customers experience them. The majority of support cases are resolved automatically without human intervention.
Scale. Trillions of telemetry metrics ingested daily from 20 billion+ sensors across HPE’s global infrastructure fleet.
Cost to operate. Petabytes of data processed continuously with efficient resource utilization through Akka’s backpressure and streaming model.
Business impact. Customer satisfaction transformed — infrastructure issues are predicted, diagnosed, and resolved before any service disruption occurs.
HPE’s predictive infrastructure management is a natural fit for agentic AI. On the Akka Agentic AI Platform, each customer deployment becomes a governed agent that continuously monitors its health, predicts failures using fleet-wide pattern recognition, and autonomously executes remediation workflows — firmware updates, configuration adjustments, capacity rebalancing — within policy guardrails.
These infrastructure agents can coordinate across the global fleet, sharing learned patterns while respecting data sovereignty boundaries. With event sourcing and durable state, every automated intervention is fully auditable, meeting enterprise compliance requirements. Enterprises build on the Akka Agentic AI Platform directly, or have a system delivered and operated through Akka Specify.
