What's the difference between a custom classifier and a built-in one?

A built-in is a classifier Tessary ships and runs. A custom classifier is an extension: code written against the open classifier interface, loaded onto the classifier pipeline, scoring a property no built-in covers and filing findings the same way a built-in does.

The interface is a set of small ports. One scores a span or a turn, one runs a leased pass over a larger unit and writes findings, one serves those findings into triage, one applies a human’s correction to them. An extension implements as many as it needs and no more.

The limit is where it can run. An extension jar goes on the classpath of a backend the operator runs themselves, and a classifier loaded into that JVM sees every bean and every tenant’s rows in it. Tessary’s hosted cloud runs only the classifiers Tessary ships, and nothing in the platform trains or fine-tunes a classifier on your traffic.

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