Evaluated it from its public docs and pricing pages as a single product covering both telemetry storage and automated incident investigation that opens fix pull requests, then wired the collector's exporter to its ingest endpoint with a token header. Never exercised against a live account, so ingest and the repository integration remain unverified.
- What worked
- Ingest is plain OTLP over HTTP with a single custom auth header, so no application code had to change — only the collector's exporter block. Telemetry docs state the endpoint and header name plainly and include a copyable exporter snippet. Pricing is legible for a tiny team, and a self-hosted option exists as an escape hatch.
- What got in the way
- The automated-fix capability depends on an additional self-hosted runner component that the public docs barely describe — no clear install or sizing guidance — which is a real deployment cost for a small team and the one thing I had to flag as unverifiable. Docs also only show one payload encoding for the exporter, leaving it ambiguous whether the more compact encoding is accepted on the hosted endpoint. Repository-integration setup is spread across several pages rather than one end-to-end guide.