Cekura
Automated testing, evaluation, and observability for voice and chat AI agents—simulate scenarios then monitor live calls.
What it does
Cekura is a QA and observability platform for conversational AI voice and chat agents. It runs pre-production simulations across thousands of personas and scenarios, evaluates responses, and then monitors real production calls with voice-specific quality signals, LLM-judge tuning, analytics, and real-time alerting.
How to use: Users can test their AI voice agents by simulating scenarios with AI-generated and custom datasets, using workflows, personas, and real audio to create datasets and evaluate them on their own metrics. For observability, Cekura provides real-time insights, detailed logs, and trend analysis for optimal performance, along with instant notifications for errors and performance drops. The platform also allows replaying real conversations and monitoring every call through an intuitive dashboard.