Grafana Labs acquires TailCtrl to accelerate development of Adaptive Traces for cost optimization

Here at Grafana Labs, we know that as our users’ IT architectures become increasingly distributed and complex, it can become increasingly difficult for them to get the visibility they need into application performance.

Distributed tracing is a powerful tool to help address that challenge, enabling you to unravel and gain deeper insights into those complex cloud native systems. With a distributed tracing backend, you can observe the flow of requests and data as they move from one service to another, so you can identify and resolve performance issues before they affect your users.

But as the volume of your tracing data increases, so can your observability costs. Meanwhile, it can become more and more difficult to discern those critical, needle-in-a-haystack traces from all the noise.

That’s why today, at ObservabilityCON 2024 in New York, we were excited to share that Grafana Labs has acquired TailCtrl, an early-stage company specializing in adaptive trace sampling. Building on the concept of tail-based sampling, adaptive trace sampling helps you automatically identify and retain your most valuable traces, so you can get the insights you need into application performance and availability, while optimizing your overall observability costs.

TailCtrl was founded by Sean Porter, a tenured entrepreneur and engineer who previously co-founded Sensu. When Sean launched TailCtrl in early 2024, he had a clear goal in mind.

“I wanted to solve this problem that I saw related to extracting value from distributed traces,” he says. “A well-instrumented set of services can produce a tremendous amount of tracing data, and not all traces are equally valuable. Traces are really this perfect amalgamation of metrics and logs — rich, structured data describing a request narrative. I wanted to turn that flood of tracing data into a stream of traces of interest. To do that, you need adaptive sampling.”

Sean, who first met the Grafana Labs leadership team at Monitorama in 2014, had a lot of experience using Grafana and Grafana Tempo, our open source, highly scalable distributed tracing backend that powers Grafana Cloud Traces.

Recognizing the synergies between Tempo and his work at TailCtrl, Sean put in a call to me and Raj Dutt, Grafana Labs CEO.

As Sean recalls: “I reached out to the team to say, ‘Hey, I’d love to figure out a go-to-market strategy around TailCtrl. This is the problem that I’m solving, and I need a first-class traces database to pair with it. What do you say?’”

Sean’s vision for TailCtrl — where “every trace stored is worthy of your attention” — includes support for the OpenTelemetry protocol (OTLP), and uses machine learning to transform enormous amounts of data into a simplified stream of traces that matter most, helping to uncover anomalies, identify incidents, and streamline root cause analysis.

We quickly realized that it aligned closely with our own vision for Adaptive Traces, a fully managed, adaptive trace sampling service we’re in the early stages of researching and building. Adaptive Traces will eventually join our growing suite of adaptive telemetry services — alongside Adaptive Metrics and Adaptive Logs — designed to help our users save time, reduce toil, and cut costs.

We look forward to working together with Sean to accelerate our journey with Adaptive Traces. Stay tuned for more!