Nicholas Leonard and Derek Caneja wanted to build AI voice agents, but when it came time to build their product, they realized that many of these voice agents had design flaws.
Some of these agents were built using no-code tools, which meant they could be shipped quickly to production, but the quality of the product was often poor. Other agents were made by companies that had the time and resources to spend months building specialized tools. “Developers and businesses needed an alternative,” Leonard told TechCrunch, adding that he and Kaneja also recognized that the future of software would be “coded, verified and optimized by coding agents.”
“These two insights and historical awareness provided the inspiration for VoiceRun,” said Leonard, the company’s CEO. Mr. Caneja is the company’s CTO.
Last year, they decided to launch VoiceRun, a platform that allows developers and coding assistants to launch and extend voice agents. Currently, many of these low-code platforms allow you to build voice agents using visual diagrams. Users click on a conversation flow and write prompts in the box to tell the agent how to behave. All of this can be difficult to manage, Leonard says.
VoiceRun, on the other hand, allows users to code how they want their voice agent to behave, giving them more flexibility to create the product they want. Leonard explained that code is a coding agent’s native language. “They’ll do a much better job working with code than with a visual interface,” Leonard said.
Additionally, visuals have limited configuration options, so if you want to build a voice agent that can speak in different dialects, for example, that can be more difficult if the maker of the visual interface hasn’t built the functionality to handle that task.
“But it’s incredibly easy to do in code,” he said. “There’s a long tail of millions of examples of little things you might want to do that aren’t supported by a visual interface.”
tech crunch event
san francisco
|
October 13-15, 2026
Apart from coding agents, VoiceRun allows users to run A/B tests and instantly deploy with one click.
The company is aimed at enterprise developers and can help businesses incorporate AI into customer service and technology companies launch voice-based products. For example, he mentioned working with a restaurant tech company to launch an AI phone concierge for dining reservations.
The company announced Wednesday the completion of a $5.5 million seed round led by Flybridge Capital.
There is a lot of competition in the AI agent space. Startups in this space took in billions of dollars last year (out of billions that flowed into AI companies generally). Leonard feels his company faces opposite ends of the market. He said there are no-code voice builders like Bland and ReTell AI that allow users to build simple demos. There are also more sophisticated tools, such as LiveKt and Pipecat, that give developers “maximum control.” He feels that Voicerun falls somewhere between these two ends.
“We deliver a global voice infrastructure and reputation-driven lifecycle while keeping ownership of business logic code and data in the hands of our customers,” he said. “The key difference is that we close the loop on end-to-end coding agent development. Developers expect to oversee coding agents that write code, run tests, deploy, and suggest improvements.”
In some ways, Leonard hopes his product will help developers create voice agent tools, which in turn will help people become more familiar with automated voices. Customers today feel “reassured” when a human answers the phone “because voice automation was weak and ineffective.”
According to a survey conducted by Five9 last year, three-quarters of survey respondents still prefer talking to a human when it comes to customer service. Leonard said he wants to change this perception because “human agency today has its own limitations,” such as language barriers and people feeling judged.
“There were great cars before the Model T, but cars never made it to the assembly line,” Leonard says. “We have great voice agents today, but they won’t take off until we build a voice agent factory. VoiceRun is that factory.”
