LoginSign Up
← Back to all sparks
A

Axiom

ANALYTICS
Velocity2.5

Log management and observability platform for events and traces.

axiom.co ↗

Axiom unifies logs, traces, and metrics into one AI-agent-queryable observability surface

observabilitymetricsai-agentsmcpevaluationstelemetry
◆Current state
Axiom is building AI-agent-native observability. Metrics reached GA unified with logs and traces and queryable by agents through MCP, Correlations stitches the three datasets together for investigations, and a run of skills (Query Metrics, Write Evaluations, SRE) turn AI coding agents into operators of the platform. Online evaluations extend it into AI-engineering workflows.
◆Where it's heading
Two arcs converge: completing the observability triad (logs + traces + metrics under one query layer) and exposing that layer to AI agents as a first-class consumer via MCP and purpose-built skills. Axiom is also moving up the stack into evaluating AI systems, not just observing infrastructure.
◆Prediction
Expect more agent-facing skills and deeper AI-engineering evaluation tooling, given the steady cadence of MCP-queryable features and eval releases across this window.

◆Recent moves

  1. 14d ago

    Correlations

    Correlations connects logs, traces, and metrics so an investigation can move from first clue to surrounding system state without manual query stitching. Builds on the now-unified triad to speed root-cause work.

    View source ↗
  2. 3mo ago

    Metrics are now generally available

    ⚡ SPARK

    Metrics reaches GA as a first-class pillar unified with logs and traces and fully queryable by AI agents through MCP and a dedicated skill. Completes the observability triad and makes the whole surface agent-operable.

    View source ↗
  3. 3mo ago

    Dashboards API

    A Dashboards API gives full programmatic control over dashboards. Useful infrastructure for codifying and templating observability views.

    View source ↗
  4. 3mo ago

    Write Evaluations skill for AI agents

    A Write Evaluations skill turns AI coding agents into authors of evaluation suites. Extends Axiom's agent-skills push into the AI-engineering workflow.

    View source ↗
  5. 4mo ago

    Online evaluations for AI engineering

    Online evaluations score AI outputs against live production traffic in real time. A meaningful step into AI-engineering observability beyond infrastructure telemetry.

    View source ↗
  6. 4mo ago

    Query metrics skill

    A Query Metrics skill turns AI coding agents into metrics-exploration experts. Part of the consistent pattern of making Axiom's surface agent-operable.

    View source ↗