LoginSign Up
← Back to all sparks
A

Axiom

ANALYTICS
Velocity0.0

Log management and observability platform for events and traces.

axiom.co ↗

Axiom completes the logs-traces-metrics triad and bets the product on AI engineering.

observabilitymetricsai-engineeringevaluationsmcpdashboards
◆Current state
Axiom has made Metrics generally available, unifying them with logs and traces and exposing all three to AI agents via MCP and dedicated skills. Around that, it is building an AI-engineering product line: online evaluations against production traffic, eval-authoring and metrics skills, and a dashboards API.
◆Where it's heading
Two reinforcing directions: completing the observability triad so Axiom is a full telemetry backend, and positioning that backend as the substrate for AI engineering, where agents both query the data and have their outputs scored against it.
◆Prediction
Expect more AI-engineering surface, deeper agent-facing skills and evaluation tooling, building on the metrics-and-evals foundation now in place.

◆Recent moves

  1. 2mo ago

    Metrics are now generally available

    ⚡ SPARK

    Metrics reach general availability, hyper-cardinality and unified with logs and traces, and queryable by AI agents through MCP. This completes Axiom's observability triad and slots directly into its AI-agent story.

    View source ↗
  2. 3mo ago

    Dashboards API

    A Dashboards API gives programmatic control over dashboard creation and management. It extends Axiom toward dashboards-as-code, consistent with its API- and agent-driven posture.

    View source ↗
  3. 3mo ago

    Write Evaluations skill for AI agents

    A Write Evaluations skill lets AI coding agents author evaluation suites. It deepens the agent-facing skill catalog around Axiom's AI-engineering line.

    View source ↗
  4. 3mo ago

    Online evaluations for AI engineering

    ⚡ SPARK

    Online evaluations score an AI capability's outputs against live production traffic in real time. This stands up evaluation-against-production as a core capability of Axiom's AI-engineering product.

    View source ↗
  5. 3mo ago

    Query metrics skill

    A Query metrics skill turns AI coding agents into metrics-exploration experts. Another entry in the agent-skill catalog, extending the MCP-queryable metrics surface to agents.

    View source ↗
  6. 4mo ago

    Field sorting by frequency, improved APL autocompletions, and better API token management

    Field sorting by frequency, smarter APL autocompletions, and improved API token management land together. A bundle of query-experience and administration refinements.

    View source ↗