A case is Tessary's unit of investigation. Classifiers produce findings, triage runs on those findings, and the ones ruled real issues turn into cases, each with a lifecycle to open, work, and resolve.
A case contains what moved, by how much, and since when, in the detector's own numbers. It carries the triage ruling in prose, with a confidence and its citations, and says which lane ruled it: from trace evidence alone, or grounded in the repository, where the citations are file paths you can open. It also holds before and after exemplars, real traces from each side of the shift, so you can see what the numbers look like in practice.
From a case, Tessary runs root cause analysis: an agentic session in a fresh E2B microVM over the failing cohort, with the trace evidence, a dossier, and the connected repo. It returns grounded hypotheses plus the checks it ruled out, and cites the files involved.
8 questions
Answered, plainly.
Two ways to run Tessary.
Tessary is an open-source agent reliability platform. Cloud and self-hosted run the same workflow on the OpenTelemetry traces your agent already emits.
Tessary Cloud
We host it for you. Send your first trace with nothing to deploy and no model key.
- traces
- 10,000 per calendar month
- stored trace data
- 1 GB
- retention
- 30 days
- model credit
- $10, one-time, for triage and root-cause analysis
- credit card
- not required
Self-hosted Tessary
Run the open-source code on your own infrastructure with one command. Add your own model key for triage and root-cause analysis.
Self-host Tessary for me by following https://github.com/tessaryai/tessary/blob/main/setup.md
docker compose -f oci://docker.io/tessaryai/tessary:compose up -d -y