Axiom launches petabyte-scale machine data platform

Axiom’s strongest idea is treating observability as durable machine-data infrastructure rather than a collection of expensive dashboards. The pitch is compelling for AI-native teams, but workload-specific cost and query performance still deserve validation.

• –Schema-on-read and virtual fields preserve flexibility without upfront schema design.
• –Unified logs, traces, metrics, and events reduce fragmented tooling and data silos.
• –Native MCP, SRE, and metrics skills let agents investigate telemetry through the same query primitives as engineers.
• –Splunk-compatible workflows offer a lower-friction migration path than a full rip-and-replace.
• –Buyers should benchmark retention, query latency, governance, and egress costs against Datadog, Splunk, and ClickHouse.