Braintrust for your voice agent
Braintrust is an AI engineering platform for prompts, datasets, experiments, evaluations, observability, and production AI apps. Use it to ship better AI systems with faster debugging, clearer evals, and tighter feedback loops.
Call 1800 570 693 to talk to the demo agent. Plans start at A$49 a month.




Integration record
- Status
- Available
- Plan
- Every plan, from Starter at A$49 a month
- How it connects
- With an API key. You create a key in Braintrust and paste it on a secure page in Connected apps.
- What it reads
- 4 actions, such as fetch Events, get Braintrust Resource and list Braintrust Resources.
- What it changes
- 13 actions, such as create Dataset, create Experiment and create Project.
- What starts an action
- The agent runs an action during a call or chat when the conversation needs it. It only runs the actions you switch on.
- Custom work
- You choose each action from its list of 17. Anything the list doesn't cover needs a custom API tool on Scale, or a Concierge build.
- Checked
- Action list checked 14 November 2026. We haven't tested each action.
Braintrust actions
Reads 4
- Fetch Events
- Get Braintrust Resource
- List Braintrust Resources
- Query Braintrust Data
Changes 11
- Create Dataset
- Create Experiment
- Create Project
- Create Prompt
- Insert Dataset Events
- Insert Experiment Events
- Summarize Experiment or Dataset
- Update Dataset
- Update Experiment
- Update Project
- Update Prompt
Deletes 2
- Delete Project
- Delete Experiment, Dataset, or Prompt
Braintrust setup
Pasting a Braintrust API key
Create an API key in Braintrust, choose Braintrust in Connected apps and paste the key on the secure page.
Choosing the actions
Choose which of the 17 actions the agent may use. Start with lookups and add the ones that change Braintrust when you're ready.
Testing, then publishing
Watch the agent use the live app in the browser tester, then publish. Voice and text agents share the connection.
More integrations
Get started
Tell us what your callers ring about, and we build and test the agent for you.


