
Modern observability has closed the visibility gap for applications. The living inventory of the underlying infrastructure — what exists, how it connects, and how it changes over time — remains a blind spot. Toise fills it: an OpenTelemetry-native, temporal graph designed so your AI assistant can answer an operator’s questions directly.
Ingests OTLP entity events from any producer — senhub-agent, OTel Collector or your own instrumentation. It speaks the standard, not a proprietary protocol.
A bi-temporal, event-sourced graph. See the current reality, or ask what you knew at 09:00 last Tuesday. Full change history is queryable.
A native MCP server lets AI assistants query topology and change in plain language — no dashboards to pivot through.
Apache 2.0, a single Go binary, no external dependencies. A GraphQL API for humans, an MCP server for assistants.