Observability, for you
Noemata gets you started with Observability and OpenTelemetry in minutes. It automatically detects applications you use (like Claude Code), collects telemetry from them, and installs dashboards — so you can start monitoring your agents, your machine, and your services.
Today, Noemata works well for:
- Keeping track of your and your team’s usage of Claude Code
- Monitoring data from your machine itself
- Application & infrastructure Observability
Built for agents, owned by humans
Noemata is a semantic layer on top of your analytical database. Frames define the entities you care about — services, tools, hosts, API endpoints, users — as views over your raw Observability data. A frame describes the schema, formatting, and aggregations in a shareable, composable way, so the right query is easy to write and stays right even as your data changes.
Noemata’s virtual file store lets human and agent knowledge become part of this semantic layer. Runbooks, dashboards, post-mortems and notebooks all live next to frames, in a versioned, semantically searchable workspace. When an agent investigates an alert it can locate the relevant runbook, pull the matching frame, and run the query — all from the same address space. Over time, the workspace becomes a shared knowledge base that speeds up every investigation, for humans and agents alike.
Documentation
- Getting started — install the CLI and open your first dashboard
- Concepts — how the file store, frames, and semantic layer fit together
- Connect your data — databases, telemetry, integrations
- Authoring Frames — design views and pages — the core authoring surface
- Build integrations — package frames, skills, and collector configuration for a data source as an npm package
- Reference — CLI commands, config, schemas