Nicholas Leonard and Derek Caneja wanted to build an AI voice agent, but while building the product, they felt that many of these voice agents had design flaws.
Some of these agents are built with codeless tools, meaning shipping to production is fast, but product quality is often poor. Other agents are created by companies that have the time and resources to spend months building custom tools. “Developers and companies need an alternative,” Leonard told TechCrunch, adding that he and Caneja also realize that the future of software will be “coded, validated, and optimized by coding agents.”
“These two historical insights and realizations inspired VoiceRun,” said Leonard, the company’s CEO. Caneja is the company’s CTO.
Last year, he decided to start VoiceRun, platform which allows developers and coding assistants to launch and scale voice agents. Today, many of these low-code platforms allow people to create voice agents with visual diagrams, where people click into a conversation flow and write directions into a box that then dictate how the agent should behave. All of that can be difficult to manage, Leonard said.
VoiceRun, on the other hand, allows users to code how the voice agent behaves, giving them more flexibility to create the product they want. Code is the native language of coding agents, Leonard explained. “They’re going to do a better job in the code than in the visual interface,” Leonard said.
Furthermore, visually, there are limited configuration options, so, for example, if someone wants to create a voice agent that can speak in different dialects, it may be more difficult if the visual interface provider does not create a feature that can handle the task.
“But in the code, it is incredibly simple to do,” he said. “There are millions of examples of little things you want to do that are not supported by the visual interface.”
Techcrunch event
San Francisco
|
13-15 October 2026
Aside from the coding agent, VoiceRun also allows users to perform A/B tests and deploy quickly with one click.
The company is geared toward enterprise developers, helping companies, for example, integrate AI into customer service, or helping tech companies launch voice-based products. He cited, for example, working with a restaurant technology company to launch an AI phone concierge for food reservations.
The company announced on Wednesday it closed a $5.5 million seed round led by Flybridge Capital.
There is a lot of competition in the AI agent space. Startups in this area last year nabbed billions of dollars (of the many billions that flooded into AI companies in general). Leonard feels his company is fighting two ends of the market: There are code-free voice builders, like Bland and ReTell AI, he says, that allow users to quickly create demos. There are also more sophisticated tools, such as LiveKt and Pipecat, which give developers “maximum control.” He felt Voicerun sat in the middle of these two extremes.
“We provide global voice infrastructure and an evaluation-driven lifecycle, while keeping ownership of business logic code and data in the hands of our customers,” he said. “The main difference is that we close the loop for the development of coding agents end-to-end. We expect developers to monitor coding agents who write code, run tests, deploy, and propose improvements.”
In some ways, Leonard hopes his product can help developers create voice agent tools that will help people become more comfortable with automated voices. Customers are now “relieved” when someone answers the phone, “because voice automation has been fragile and ineffective.”
Survey from Five9 last year shows three-fourths of the survey respondents still like to talk to people when it comes to customer service. Leonard says he wants to change this perception because “human agents now have their own limitations,” like language barriers or making people feel judged.
“There were big cars before the Model T, but vehicles didn’t go anywhere until the assembly line,” Leonard said. “There are great voice agents out there today, but they won’t be anywhere until the voice agent factory is built. VoiceRun is that factory.”