Our investment in Vapi: the voice AI developer platform

Voice AI isn’t just an upgrade to software’s UI. It’s transforming how businesses and customers connect.

Despite telephony being an over 150 year-old technology, voice still remains the dominant mode of communication for most businesses. In large industries like healthcare, legal services, home services, insurance, and logistics, businesses rely on phone-based communication to convey nuanced information more effectively, provide more personalized support or services, handle critical transactions, and address time-sensitive needs. 

We’re in the midst of a Cambrian explosion of voice AI infrastructure that has driven our conviction in the market opportunity in this space. Developers are in a better position than ever to integrate AI voice capabilities into their products and services to help eliminate long wait times, automatically resolve customer issues, and scale voice agents infinitely to meet spikes in customer demand any time of day or night. That said, building, deploying, and managing conversational voice agents at scale is a substantial challenge. 

This is why we’re thrilled to lead Vapi’s $20 million Series A financing. Vapi is building the leading developer platform for conversational voice agents, abstracting away the complexity of building these agents and managing real-time infrastructure, and making it easy for anyone to deploy human-level voice experiences.

Vapi provides a 10x improvement on the development experience for voice agents. Its elegant platform solves the hard engineering problems required to manage real-time voice infrastructure and build enterprise-grade voice agents, including the orchestration of third-party speech-to-text (STT) models, LLMs, and text-to-speech (TTS) models.

Vapi’s architecture is carefully designed to help developers create agents that sound lifelike without sacrificing the control and determinism that enterprise customers demand. From its proprietary real-time audio model that can discern speakers’ emotional inflections, to bulletproof scalability, reliability, and fault tolerance, no detail has gone untested. This level of detail stands out because it addresses a core enterprise AI problem: AI models are probabilistic in nature, yet highly regulated industries—think financial services, healthcare, and insurance—require deterministic output. Vapi solves this conundrum by breaking conversational flows into digestible, discrete blocks.

Our conviction to lead Vapi’s Series A came especially from our strong belief in its team. In a space that is evolving as fast as AI, we believe that execution, customer obsession and product velocity matter far more today than ever before. 

We’ve been impressed by how quickly cofounders Jordan Dearsley and Nikhil Gupta have built a robust platform with a vibrant community of enthusiastic developers, all with a lean and agile team. The two Waterloo engineers have iterated extremely quickly—it’s not uncommon for us to hear that a feature request made on a Friday evening is in production on Saturday morning. 

Vapi is rapidly growing, and customers praise its ergonomic developer experience, extensibility, speed-to-execution, and reliability. 

We’re thrilled to partner with Jordan, Nikhil, and the rest of the Vapi team as they enable developers to dream bigger about the range of voice experiences they can power.