Can $156M Help Observe Redefine Observability for the AI Era?

The gold rush for AI has created an invisible side effect: an explosion of telemetry data that threatens to overwhelm the very systems built to keep applications running smoothly. Every new service and model generates more logs, traces, and metrics, turning observability from a quiet backend function into one of the most expensive line items in modern IT.

Observe believes it can change that. The San Mateo startup just raised $156M in Series C funding to scale a platform built on a telemetry data lake, a real-time knowledge graph, and AI-powered SRE. Its goal is to give engineering teams faster answers and lower bills, taking on more established players in the market like Splunk, Datadog, and Elasticsearch in the process.

A major obstacle for data teams is how fragmented the information becomes after it is collected. Telemetry often ends up spread across different tools, forcing teams to piece together a complete view during outages or performance issues.

The cost side is just as challenging. Legacy pricing models that bill by ingestion or storage can send costs soaring as AI workloads expand. For many organizations, observability has shifted from an operational safeguard to a budget headache, and that is the gap Observe is aiming to close.

“During a period of explosive growth at Tekion, we realized our existing observability tools weren’t going to scale with us,” said Binu Mathew, CTO at Tekion. “We had tried major commercial and open-source tools, but both resulted in escalating costs and constant tuning efforts that drained engineering resources. Observe gave us a cost-effective unified platform for logs, metrics, and traces, with the ability to correlate across all of them.”

According to the company, its platform is built to pull logs, metrics, and traces into a single Telemetry Data Lake. Data is ingested in real time and stored in an open, compressed format that, they say, keeps storage costs predictable as workloads scale. This design also removes the heavy indexing and constant tuning, they argue is common in older systems.

Observe says the next layer, its live Knowledge Graph, maps how services, infrastructure, deployments, and incidents are connected. That context, the company claims, allows data teams to skip the manual stitching often required during outages.

On top of that sits the AI SRE, which the company describes as an always-on system for spotting anomalies, flagging root causes, and recommending or triggering fixes. Together, these three components are meant to speed up troubleshooting while easing the operational burden of managing observability at scale.

“Our customers rely on us to unify data from hundreds of sources, which demands a highly scalable and efficient infrastructure,” said Andrew Katz, CTO and Co-Founder at mParticle. “Observe’s data lake-based architecture allows us to scale observability much more easily and cost-effectively than traditional solutions.”

The observability market is dominated by long-established platforms, but many of them rely on designs that can be hard to manage at today’s scale. As AI workloads multiply, organizations are running into the limits of architectures that demand constant indexing, tuning, and storage oversight.

This has created an opening for platforms that can simplify operations while keeping costs in check. Observe is looking to step into that space, promoting a model it says can handle modern data demands without the same maintenance burden.

Large-scale rollouts show how the platform is being put to work. Telemetry is stored in Apache Iceberg format, allowing customers to keep full control of their data and avoid lock-in. The system also uses OpenTelemetry for collection, making it easier to plug into existing pipelines and tooling.

Earlier this month, Observe added an MCP server that lets external AI SREs work directly with its observability context. This opens the door for partners and even other tools to take part in automated incident workflows powered by the same real-time knowledge graph.

The company claims that a major international bank replaced Splunk with Observe to process 30TiB of compliance logs per day, later scaling to nearly 100TiB with more than 3,000 users. Splunk was retired entirely, and the bank now plans to phase out AppDynamics in favor of an OpenTelemetry-native APM strategy.

Observo also shared numbers that suggest that their approach is resonating with some customers. Over the past year, the company has tripled its revenue, doubled the number of enterprise customers, and is now handling more than 150PB of data.

Knowledge Graph (Source: Observe)

It also reports a net revenue retention rate of 180%, meaning existing customers are using the platform more over time. Among them are Topgolf, which uses Observe to keep ingestion costs tied directly to resource usage, and Dialpad, which says the platform has cut troubleshooting time by as much as 30%.

For Capital One Ventures, the real draw is how Observe positions itself at the heart of system reliability. Partner Sean Leach called full-stack observability “foundational for AI” and a critical layer for tracking resource use and delivering tailored customer experiences. He said the firm is backing Observe because it is “executing on a bold vision for modern observability,” and they want to help accelerate that progress.

Snowflake Ventures is also deepening its commitment to Observe. It noted that a telemetry-first design could be a natural fit alongside the Snowflake Data Cloud, creating opportunities for joint solutions in enterprise environments.

Earlier this year, Observo AI made headlines with a $15M seed round to launch its agentic AI-powered data pipeline platform. Now, Observe has grabbed the spotlight with a massive $156M Series C, a sign that investors see real potential in AI-led solutions for the observability crunch. With this new funding, Observe has the runway to grow its platform, add new features, and push harder against rivals.

ScaleOut Enhances Digital Twin Intelligence With Generative AI and ML

ETL vs ELT for Telemetry Data: Technical Approaches and Practical Tradeoffs

The post Can $156M Help Observe Redefine Observability for the AI Era? appeared first on BigDATAwire.