Skip to content
How we review
ai-agent.review — AI Agent Reviews, home
← All news

Product updates

Ada lets support teams score conversations against their own criteria.

Vendor-reportedChecked: [release][docs]

Editorial attribution: Gosia Madej

Prepared with AI assistance. Editorial attribution does not imply personal product testing or manual review of every story.

Published: · Event date:

Ada’s November 16 release pairs custom quality metrics with a workflow that can propose and test fixes before a team approves rollout.

Vendor-reportedChecked: [release][docs]

What changes for users

Have reviewers score a sample before using the metric as an adoption gate. If a pass rate shifts, check for changes in the criterion or judge as well as changes in the agent’s behavior.

Editorial assessmentChecked: [release][docs]

A red check mark on a paper proof beside a clear ruler.
AI-generated editorial illustration. Not a product screenshot. · AI Agent Reviews (AI-generated editorial illustration)

Ada says both features are available on every account. Operators write criteria in plain language and review the judge’s verdict, explanation and supporting quote. A metric can surface a recurring failure, such as the agent asking twice for information a customer already supplied.

Vendor-reportedChecked: [release][docs]

An assistant using Ada’s MCP tools or CLI workflow can group failures, draft a change set and run up to three rounds of simulations. Ada says the assistant asks before creating a test, running each test and rolling out or promoting a change.

Vendor-reportedChecked: [release][docs]

Checked against sources

AI-assisted source check

How this was checked

We compared Ada’s dated release note with its Custom Metrics documentation.

Editorial assessmentChecked: [release][docs]

We did not run tests in Ada or validate the judge’s accuracy.

Editorial assessmentChecked: [release][docs]

Source evidence

Original source excerpt. This shows what the source says; it does not establish product performance.

Custom Metrics and the improvement loop are available on every account.

Read original source [release]

Sources