Cekura

"Automated QA for Voice AI and Chat AI Agents"

Cekura is a testing, monitoring, and continuous-improvement platform for conversational AI agents. It helps teams simulate thousands of voice and chat conversations before release, evaluate agent behavior, monitor production calls, detect regressions and security failures, and turn real-world issues into repeatable tests. Launched in 2024, it now works with more than 150 conversational AI companies, including over 25 enterprises.

Product Details

Other details

Category : Testing

Pricing : Freemium

Stage : Active Customers

Customers : Deloitte, PwC, Jobber, Confido Health, Kastle, Twin Health

Makers:

Sidhant Kabra , Tarush Agarwal , Shashij Gupta

Key Features

Pre-production simulation of thousands of synthetic voice and chat conversations
Custom evaluations for accuracy, latency, compliance, hallucinations, and workflow completion
Production monitoring with call analytics, tracing, alerts, and drift detection
Adversarial red-teaming for jailbreaks, data leaks, toxicity, and off-script behavior
Self-improving loops that reproduce failures, suggest fixes, and rerun release gates
Side-by-side benchmarking across agent platforms, models, and infrastructure providers

Headquarter At:

San Francisco, USA

Bengaluru, India

Lifecycle

Rebranded

Product status: Active

12 March 2025 · New name: Cekura

Vocera rebranded as Cekura to reflect its focus on testing and monitoring conversational AI agents.

View announcement