Dstl8 distills runtime signals at the source and routes diagnoses back to the agent or engineer who shipped the code.

ControlTheory today announced the general availability of Dstl8, its runtime feedback loop platform for AI-speed engineering. Where observability tells teams what already happened, Dstl8 closes the loop back into development itself: it distills runtime signal at the source and routes the answer back to the agent or engineer who shipped the code. Dstl8 is the runtime feedback loop for AI-speed engineering, built to keep pace.

AI has dramatically changed development speed, while monitoring approaches have stayed the same. Engineering teams are shipping more code, faster, while most teams' observability tooling was built for a slower, more predictable cadence of human-generated code and dashboards. The issue compounds: more volume, more velocity, and all at the same time. The sheer amount of code moving through production is outpacing the ability of a traditional observability architecture to monitor what's happening and feed the right answers quickly back to the right team.

Telemetry Distillation: How ControlTheory’s Dstl8 Platform Works

Dstl8's mechanism is Telemetry Distillation: it runs locally where the data starts, at scale, and extracts what matters before anything ships or gets piped into an expensive observability platform. Dstl8 follows four steps, in order:

1. Distill. Reads the actual content of every line of telemetry, not just its status code, across four signal dimensions: sentiment, patterns, anomalies, and severity.
2. Correlate. ControlTheory's custom, fine-tuned model connects root cause across the full deployment chain, keyed to the deploy event.
3. Reason. Delivers diagnoses with cited evidence directly into the tools engineers already use through MCP, including Claude Code, Cursor, and Codex.
4. Remember. Every triage feeds a knowledge graph, so past incidents and prior fixes surface automatically on the next one.

Sentiment is a new signal dimension for the industry. Dstl8 reads the content, tone, and confidence of what a system is saying about itself, catching degradations that may have never tripped a standard threshold.

Dstl8 works across Kubernetes, AWS CloudWatch, OpenTelemetry, Google Cloud, Supabase, Vercel, Railway, and more. Infrastructure and configuration problems are pinpointed with a recommended fix; application code problems get a diagnosis down to the line of code. Every investigation, fix, and resolution becomes institutional memory.

"Code volume and code velocity are both climbing at once, and they amplify one another. The real problem isn't necessarily bad code; it’s just more of it, moving faster than any team can keep up with and continuously feed back to the team that can fix it. Observability without feedback is just watching. A feedback loop tells the agent what to do next," said Bob Quillin, CEO and co-founder of ControlTheory.

Proof: Zero Alert Rules

An early ControlTheory customer in enterprise fintech running Dstl8 across 13 Kubernetes clusters over a two-month period automatically surfaced and resolved 328 incidents, without a single alert rule written by their team.

Dstl8 builds on ControlTheory's open-source terminal tool, Gonzo, an MIT-licensed TUI (terminal UI) for real-time log analysis that has drawn more than 2,700 GitHub stars since its release last year. Where Gonzo is a local, single-session tool, Dstl8 is continuous, shared across a team, and built for the enterprise.

Availability

Dstl8 is generally available today. Teams can get started at Dstl8.ai. Company leaders will be providing in-person demos at KubeCon + CloudNativeCon North America in Salt Lake City, November 9-12, 2026.

About ControlTheory

Founded in 2024 by observability veterans from StackEngine (acquired by Oracle Cloud) and CopperEgg (acquired by Idera), ControlTheory builds runtime feedback tools for AI-speed engineering without the cost and complexity of traditional observability systems. The company's Dstl8 platform applies Telemetry Distillation to surface sentiment, patterns, anomalies, and severity at the source, then reasons across topology, code graphs, and runtime signals to pinpoint problems in staging and production, with zero alert rules to write. Learn more at controltheory.com.

View source version on businesswire.com: https://www.businesswire.com/news/home/20260922140469/en/

Observability without feedback is just watching. A feedback loop tells the agent what to do next.