At KubeCon+CloudNativeCon London, Bob Quillin, CEO and Co-Founder of ControlTheory, sat down with us to discuss how his startup is addressing the mounting frustrations around observability costs and effectiveness. The newly-launched company has secured $5 million in seed funding to build an intelligent control plane that sits atop existing observability stacks, promising to tame runaway costs while improving signal quality. In this exclusive interview, Quillin explains why, despite years of innovation, observability remains fundamentally “broken” and how ControlTheory’s approach could finally deliver on its promises.

The Genesis of ControlTheory

Despite advancements in observability tooling, many fundamental problems remain unsolved. “We’ve been in the observability space most of our careers,” explained Quillin. “About two years ago, we kept wondering why it’s so broken. There’s a lot of frustration. Customers feel they’re paying too much, costs are spiraling out of control, telemetry volumes are going up, and complexity is high.”

The ControlTheory team identified several critical issues: costs continue to rise, root cause analysis hasn’t significantly improved, business KPIs remain difficult to extract, and mean time to repair (MTTR) is still increasing. As Quillin noted, “All the things you’ve got observability for are broken.”

This realization led the team to explore a new approach focused on feedback loops and intelligence. Rather than the traditional “one-way pipes that dump a lot of data” where users pay for ingestion, indexing, and retention of everything, ControlTheory aims to deliver “the data that you want, when you need it and for whom” by being more intelligent about data delivery without requiring changes to existing systems.

A Proven Team with Deep Experience

ControlTheory brings together a leadership team with extensive experience working together across multiple successful startups:

• Bob Quillin, CEO
• Eric Anderson, CTO
• Robert Gordon, Chief Architect
• John Reeve, Chief Product Officer

The team previously founded StackEngine, a container management and Kubernetes management service acquired by Oracle Cloud, where they built out Oracle’s managed Kubernetes service. Before that, many team members worked together at CopperEgg (a cloud monitoring company acquired by Idera that competed directly with early DataDog), and Hyper9 (virtualization management).

“We have a long history of working together as a team,” said Quillin. “It’s great to know who you’re working with.”

ControlTheory recently secured $5 million in seed funding from Silverton Partners, which has also funded the team’s previous ventures.

The ControlTheory Approach

ControlTheory breaks down its approach into three key areas: cost control, operational control, and adaptive control. The company’s solution sits as a control plane on top of the OpenTelemetry Collector, connecting with any source and destination.

For cost control, the platform addresses common challenges like log spikes by:

• Filtering unnecessary logs before they’re sent to observability vendors
• Deduplicating redundant logs to avoid paying for ingestion and indexing
• Using “meta metrics” to provide “observability of your observability” to identify where excessive logs originate
• Reducing cardinality in custom metrics, which can reduce costs by 20-30%

On the operational side, ControlTheory helps improve signal-to-noise ratio, ensuring teams get the information they need when they need it. The platform can also support AI observability solutions by providing refined, specific data to these systems.

For adaptive control, the company has developed “elastic telemetry pipelines” that automatically tune themselves based on feedback loops, scaling up when more information is needed and scaling down when it’s not.

Real-World Use Cases

At KubeCon, Quillin highlighted several compelling use cases that demonstrate the need for ControlTheory’s approach:

1. Privacy and security compliance: Organizations handling sensitive information like credit card data (e-commerce) or protected health information (healthcare) need to mask and redact data before sending it to observability systems. ControlTheory enables them to maintain governance and compliance while still benefiting from observability tools.
2. Trace optimization: As more organizations adopt distributed tracing, they face significant cost increases due to the volume of data. ControlTheory provides tail sampling capabilities that send only the most important traces to observability solutions, improving root cause analysis while controlling costs.

“The light bulbs go off when you talk to folks about their current issues, whether they’re using Prometheus, Grafana, Splunk, DataDog, Dynatrace, or New Relic,” Quillin observed. “Across the board, these issues are pervasive.”

What’s the future

Having just launched the company and announced their $5 million funding round at KubeCon, ControlTheory is now focused on building its customer base through an early availability program. The company plans to work closely with both customers and the broader partner ecosystem, including the OpenTelemetry community and major observability vendors.

“We want to help DataDog customers make their environment better. We can help them adopt more solutions and bring in more data with higher fidelity,” explained Quillin. “We’ve  stagnated on the old problems, and now we can start solving some new problems and take that next huge leap forward.”

For SREs looking to address observability challenges in their environments, ControlTheory offers a free trial at controltheory.com.

Guest: Bob Quillin
Company: ControlTheory
Show: KubeStruck