Ignite AI: The Future of Voice AI Testing and Self-Improving Agents with Tarush Agarwal | Ep284

A multi-turn prompt attack convinced one of the biggest providers’ voice agents to issue a $150 discount code.

Tarush Agarwal is the co-founder and CEO of Cekura.ai, which builds testing and verification infrastructure for voice agents. Before founding Cekura, he studied computer science at IIT Bombay and worked on low-latency quantitative trading systems in London and Chicago, where teams optimized performance at roughly seven to nine nanoseconds. Cekura entered Y Combinator after pivoting from a legal voice-agent product, later raised a $2.5 million seed round, and now works with more than 200 customers while running millions of simulations.

Voice agents can perform well in controlled tests and still fail during real conversations. Interruptions, background noise, mixed languages, transcription errors, emotional callers, and multi-turn manipulation can expose problems that never appear in a text-based evaluation.

Tarush’s most counterintuitive claim is that better models do not automatically produce better voice agents. Around half of Cekura’s customers still use GPT-4.1 because newer reasoning-heavy models can introduce delays that do not work in live calls. Production performance depends on the full system, including latency, transcription, turn detection, interruption handling, speech quality, instruction following, and the infrastructure connecting each component.

In Today’s Episode We Discuss:

00:01 Introducing Tarush Agarwal and Cekura.ai

00:27 From IIT Bombay to quantitative trading

02:45 Founder life versus low-latency engineering

04:26 Building voice agents for personal injury law firms

06:27 Pivoting during the first week of Y Combinator

07:11 Early growth, the $2.5 million seed round, and customer focus

09:49 The current state of voice AI

12:49 The metrics that determine voice-agent quality

15:17 Compliance, healthcare, and high-stakes conversations

18:02 How multi-turn prompt attacks exploit voice agents

19:17 The quiet problem with how companies run evals

22:33 Why testing voice agents through text is insufficient

24:06 Cascading systems versus speech-to-speech models

25:36 Building realistic simulation environments

27:12 What changed in voice AI over two years

29:29 Public benchmarks, latency gains, and accuracy limits

31:14 Cekura’s long-term vision beyond voice

32:16 Moving from founder-led sales to a dedicated GTM team

33:57 The product metric Tarush watches every day

35:37 Why voice AI could become larger than software

Cekura began after Tarush and his co-founders spent three hours after dinner manually calling their own legal voice agent. He explains why healthcare teams must simulate distressed patients, how multi-turn testing exposed the $150 discount exploit, and why his team sometimes shipped a bug fix before the customer reporting it had finished the call.

The episode returns to an old engineering principle: reliability begins when reality is allowed to break the system.

Pull Quotes

“Everyone talks about evals. I don’t think most people know how to do it correctly.”