Cekura, which operates as Tatva Labs and was previously known as Vocera, builds the testing and observability layer that AI voice and chat agents need to be trusted in production. As conversational agents proliferate across support, sales, and operations, teams quickly hit the same wall: it is hard to know whether an agent will behave correctly across the messy reality of real conversations. Cekura's platform exists to close that gap with systematic simulation and evaluation.
The product lets teams define and run automated test scenarios against their voice and chat agents at scale, simulating callers and conversational paths to surface failures, hallucinations, broken tool calls, and regressions before customers encounter them. Beyond pre-launch testing, Cekura provides observability and monitoring for agents in production, so teams can track quality over time, detect drift after prompt or model changes, and respond before issues compound. This pre-launch-plus-production approach mirrors how mature software teams treat reliability for any critical system.
Founded in 2024 by Tarush Agarwal (CEO), Shashij Gupta (CTO), and Sidhant Kabra (CBO), Cekura went through Y Combinator and targets the growing population of companies building serious, customer-facing conversational agents. Its value proposition is straightforward: ship voice and chat agents with the same testing rigor expected of production software, rather than relying on manual spot-checks.
In July 2025 the company (as Cekura/Vocera) raised a $2.4M seed round led by returning investor Y Combinator, with participation from Flex Capital, Pioneer Fund, Decacorn Capital, new investor Hike Ventures, and individual angels. The funding supports building out its conversational-AI testing and observability platform as voice agents move further into production use.