VoiceRun nabs $5.5M to build voice agent factory

Nicholas Leonard and Derek Caneja embarked on a mission to create advanced AI voice agents. However, during their initial development process, they identified significant design shortcomings in existing solutions. Many voice agents were constructed using no-code tools, which enabled rapid deployment but often resulted in subpar quality. Conversely, some were developed by companies with ample resources, taking months to create specialized tools. "Developers and enterprises needed an alternative," Leonard shared with TechCrunch. He emphasized that he and Caneja recognized a pivotal shift in software's future—one that would rely on coding agents for validation and optimization. Motivated by these insights, they founded VoiceRun, a platform designed to empower developers and coding assistants to create and scale voice agents efficiently. Unlike many low-code platforms that utilize visual diagrams for agent creation, which can be cumbersome, VoiceRun allows users to write code to dictate agent behavior, offering greater flexibility. Leonard explained, "Code is the native language of coding agents. They excel in coding environments rather than visual interfaces." He noted that visual tools often limit configuration options, making it challenging to implement features like dialect variations without specific support from the interface. In contrast, coding makes such customizations straightforward. VoiceRun not only facilitates coding but also enables users to conduct A/B testing and deploy their agents instantly with a single click. The platform primarily targets enterprise developers, assisting companies in integrating AI into customer service and launching voice-driven products. One of their notable collaborations includes a restaurant-tech firm developing an AI phone concierge for reservations. On Wednesday, VoiceRun announced that it had successfully raised $5.5 million in a seed funding round led by Flybridge Capital. The competition in the AI voice agent sector is fierce, with startups securing billions in funding over the past year. Leonard views his company as positioned between no-code voice builders, such as Bland and ReTell AI, and more advanced tools like LiveKt and Pipecat. "We provide global voice infrastructure and an evaluation-driven lifecycle, ensuring that our customers retain ownership of their business logic code and data," Leonard stated. He aims to create a complete ecosystem for coding agent development, allowing developers to oversee agents that write code, conduct tests, deploy, and suggest enhancements. Leonard envisions that his solutions will help foster user comfort with automated voices. Currently, many customers express relief when speaking to a human, as they often find existing voice automation lacking. A survey from Five9 revealed that 75% of respondents prefer human interaction for customer service. Leonard aims to shift this mindset, acknowledging that human agents also face limitations. He likened the evolution of voice agents to the automobile industry's history, stating, "There were great cars before the Model T, but vehicles became widespread only after the assembly line. Similarly, while there are effective voice agents today, they will not become commonplace until we establish a voice agent factory. VoiceRun is that factory."