Nicholas Leonard and Derek Caneja desired to develop AI voice agents, but after they began developing the product, they felt that a lot of these voice agents had design flaws.
Some of those agents were built using no-code tools, which meant shipping into production was quick, but the standard of the product was often low. Other agents were made by firms that had the time and resources to spend months developing specialized tools. “Developers and firms needed another,” Leonard told TechCrunch, adding that he and Caneja also recognized that the long run of software could be “coded, validated and optimized by coding agents.”
“These two insights and a historical insight gave us the inspiration for VoiceRun,” said Leonard, the corporate’s CEO. Caneja is the corporate's CTO.
Last 12 months they decided to provide it a go VoiceRun, a platform This allows developers and programming assistants to launch and scale voice agents. Currently, a lot of these low-code platforms enable the creation of voice agents with visual graphs, where people click through conversation flows and write prompts in fields that then dictate how the agent should behave. All of that could be difficult to administer, Leonard said.
VoiceRun, then again, allows users to program how they need their voice agents to behave, giving them more flexibility in constructing the product they need. Code is the native language of programming agents, Leonard explained. “You will do a much better job in code than in a visible interface,” Leonard said.
Additionally, visual elements have limited configuration options. So, for instance, if someone desires to create a voice agent that may speak in a unique dialect, this may be tougher if the visual interface manufacturer has not developed a feature that may handle this task.
“But it’s incredibly easy in code,” he said. “There are countless examples of little things it is advisable to do this aren't supported by the visual interface.”
Techcrunch event
San Francisco
|
thirteenth–fifteenth October 2026
In addition to coding agents, VoiceRun also allows users to perform A/B testing and fast deployment with one click.
The company is geared toward business developers and helps firms, for instance, integrate AI into their customer support or help technology firms introduce voice-based products. For example, he mentioned working with a restaurant tech company that launched an AI phone concierge for meal reservations.
The company announced Wednesday the closing of a $5.5 million seed round led by Flybridge Capital.
There is plenty of competition in the sphere of AI agents. Startups on this space have raised billions of dollars previously 12 months (out of the numerous billions which have flowed into AI firms generally). Leonard believes his company is scuffling with two ends of the market: There are no-code voice builders like Bland and ReTell AI that allow users to quickly create demos. There are also more sophisticated tools like LiveKt and Pipecat that give developers “maximum control.” He seems like Voicerun sits in the midst of those two ends.
“We provide a worldwide voice infrastructure and a review-driven lifecycle while keeping ownership of business logic code and data within the hands of the client,” he said. “The key difference is that we’re closing the loop on developing end-to-end coding agents. We expect developers to oversee coding agents who write code, run tests, deploy, and suggest improvements.”
In a way, Leonard hopes his product will help developers create voice agent tools, which in turn will help people feel more comfortable with automated voices. Today, customers are “relieved” when a human answers the phone, “because voice automation was fragile and ineffective.”
A survey by Five9 shown last 12 months that three-quarters of survey respondents still prefer to talk to a human relating to customer support issues. Leonard said he wanted to vary that perception because “human agents today have their very own limitations,” akin to language barriers or the sensation that folks are being judged.
“There were great cars before the Model T, however it wasn’t until the assembly line that vehicles became ubiquitous,” Leonard said. “There are great voice agents today, but they won't be ubiquitous until the voice agent factory is built. VoiceRun is that factory.”
