What is automated testing for AI voice agents?
Automated testing for AI voice agents involves evaluating their performance, accuracy, and responsiveness using predefined scripts and test cases without human intervention. It ensures the agent correctly understands spoken input, responds appropriately, and handles various scenarios like interruptions or errors.
Which voice agent platforms does Relyable integrate with?
Relyable offers native integrations with leading platforms such as Vapi, Retell, and ElevenLabs, allowing for quick setup using an API key. For custom-built agents on other platforms, you can upload your prompt and route calls through Relyable's API for real-time monitoring.
How are calls evaluated by Relyable?
Calls are evaluated against customizable test cases, which can be defined by the user or generated from their system prompt. Each interaction is automatically analyzed using built-in metrics like latency, sentiment, and customer satisfaction, allowing for precise and consistent assessment across scenarios.
What is the pricing model for Relyable?
Relyable offers paid plans based on credits, covering both simulated calls and live call evaluations. The Core plan starts at $500/month for 50,000 credits, while the Premium plan is $1,000/month for 125,000 credits. Enterprise plans with custom credit packages are also available.
What kind of alerts does Relyable provide for live monitoring?
Relyable provides real-time alerts when a live call's performance drifts from set thresholds. These alerts can be routed to various channels including Slack, Email, PagerDuty, Opsgenie, Linear, and custom webhooks, ensuring prompt notification of issues.