Does one frustrated message trigger an alert?

No: one flagged message writes a detection, not an alert, and a case opens only when a call site’s rate of frustrated conversations rises above the rate it learned as its own normal.

The unit is the conversation, not the message. A conversation counts as frustrated once one of its messages is flagged, and scoring of that conversation stops there, so a user who complains five times still counts once. That’s what keeps one unhappy person from looking like a trend.

Each call site learns its own normal rate from its own traffic and is compared with its own past, not with a shipped number. When its rate rises above that normal, the finding and the case open together. There’s no triage step in between, because the rate on the finding is the claim. The case is what alert rules route to people.

The limit is the baseline itself: a call site that frustrates users from its first day learns that as normal, and is flagged only if it gets worse.

keep reading

More on this.

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.

what's includedper organization
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