1What it does
Use HighLevel's Prompt Optimizer to run five AI-to-AI test calls, review the failures and refine the prompt before a client ever hears the agent.
2Who it is for
- Industry
- Agency
- Platform
- HighLevel
- Category
- Voice AI
- Tools used
- HighLevel Voice AIPrompt OptimizerElevenLabs
3How it works
Every step, in order, is in Workflow at a glance, pinned on the right as you scroll.
- 01 Trigger
Agent ready for testing.
- 02 Action
Pick voice and ElevenLabs model (Flash vs Flash 2 etc.) for cost vs speed.
- 03 AI / action
Pick the GPT model for cost per minute.
- 04 Action
Prompt Optimizer: run 5 scenarios.
- 05 Action
Fix failed conversations and regenerate the prompt.
- 06 Action
Test refund, policy and edge cases.
- Trigger
- Agent ready for testing.
- AI step
- Scenario tests such as refund requests and policy questions.
4How it went
Unknown. The source shows the build but does not report results.
Only what the source reports. The source did not report numbers.
5Worth knowing
- Model choice changes cost per minute. Check before you quote a client.
6Recipe
Adapted from source Reworded or reconstructed from what the source shows.
5 test scenarios: 1) asks for a refund 2) asks a question not in the knowledge base 3) wants to book but gives vague info 4) angry caller 5) asks if it is a robot.