Startup Cekura has announced Tuesday that it raised $2.4 million in funding to make “every AI conversation reliable.” Founded by IIT Bombay alumni Sidhant Kabra, Shashij Gupta, and Tarush Agarwal, this startup provides a testing and observability layer for Voice and Chat AI agents.
Cekura’s technology simulates thousands of synthetic calls or chats in minutes, pushes agents through tricky edge cases, and monitors production conversations, catching failures before end-users ever notice.
The startup, which was founded in 2024, already safeguards voice and chat interactions for large enterprises and hyper-growth startups in healthcare, BFSI, retail, and more.
READ: Blue Onion raises $10 million in funding (April 18, 2025)
Investors who contributed to this funding round include Y Combinator, Flex Capital, Hike Ventures, Kulveer Taggar, Pioneer Fund, Chris Smoak, Ooshma Garg, Richard Aberman, JJ Fliegelman, Decacorn, Alan Rutledge, Imagination Capital, Allport Capital, Ignite Ventures among others.
Kabra told Business Insider that they were working on AI agents in the healthcare space and struggling with quality assurance (QA) amid manual fixes that could take hours. Cekura uses AI to simulate conversations and generate thousands of edge-case scenarios to put AI agents through their paces before going live.
“The customers will interrupt you, the customers will be toxic, the customers will try to jailbreak you, the customers will operate out of bias,” Kabra said. “You need to really stress test your agents before you go live.”
Cekura sees a growth in opportunities as call centers shift to AI. The company monetizes via a subscription model for startup clients, beginning at $1,000 per month. It also has custom enterprise offerings. Kabra also mentioned that while 90% of its business is focused on voice agents, it also builds chat agents.
READ: Further AI raises $5 million to build AI assistants for insurance companies (April 4, 2025)
Cekura says the funding will go towards enhancing Cekura’s scenario-generation engine, deepen coverage metrics. The company is also working on expanding to more enterprise teams and helping them get a step closer to their vision in making every conversational AI agent reliable and secure.
Cekura has roughly 70 customers across industries — including in highly regulated spaces like healthcare and financial services, where “the threshold of reliability is pretty high,” Kabra said. These clients include AI mortgage servicing startup Kastle and Sandra — an AI receptionist for car dealerships. Cekura faces competition from other startups including fellow Y Combinator grads like Coval and Hamming.
Cekura said its long term goal is to “elevate the quality bar for all AI-driven conversations.”
“We want a world where interacting with an AI agent feels helpful and secure, no matter if you’re chatting with a retail customer support bot or talking to your car’s voice assistant,” the company said.