Causely

Causely (founded 2022 by Shmuel Kliger — founder of Turbonomic, acquired by IBM, and former CTO of SMARTS, acquired by EMC) raised $8.8M in seed funding for what it calls 'causal AI': instead of correlating anomalies, it maintains a real-time Bayesian network of semantically abstracted telemetry and matches observed symptoms against a finite set of known failure modes to pinpoint root causes deterministically. It installs in Kubernetes environments in under a minute, auto-discovers service topology and dependencies, ingests OpenTelemetry streams (plus eBPF-based auto-instrumentation for teams without OTel), and surfaces only the critical risks to reliability — with automatic impact/blast-radius analysis and integrations like IBM Instana. An MCP server exposes causal context to AI agents.

The honest picture: the causal-inference framing is a substantive alternative to both dashboard-spelunking and LLM-guessing — deterministic, explainable, and cheaper in tokens than agent-based RCA. But it is Kubernetes-centric and early-stage; breadth of integrations and long-term reliability data are thinner than incumbents. Worth watching precisely because it disagrees with the 'more data + chatbot' consensus.

Deploy the Causely agent into your Kubernetes cluster (under a minute per the vendor) or connect your OpenTelemetry collector / IBM Instana via API. Topology mapping and causal analysis start automatically.