Amazon Ring needed help. The smart doorbell company was drowning in customer support calls during last year’s holiday season and faced a critical decision: expand expensive call center capacity, rely on clunky automated phone systems, or try something new with AI voice agents.
Ring evaluated more than 40 AI voice vendors before settling on startup Vapi. Today, Ring routes 100% of its inbound calls through Vapi’s platform – a deployment that helped the two-year-old company raise $50 million at a $500 million valuation.
The deal highlights how quickly enterprises are moving real customer interactions onto AI systems, moving beyond simple chatbots to voice agents that can handle complex conversations. For Vapi, founded by University of Waterloo classmates Jordan Dearsley and Nikhil Gupta, the Ring partnership validates their approach of building infrastructure rather than pre-packaged applications.
Ring turned to Vapi in mid-Q4 last year when the company was weighing its options for handling increased call volume. Dearsley believes Ring chose Vapi because it gave Ring engineers granular control over how AI agents behaved during live customer interactions – something many competitors couldn’t offer.
“A lot of AI tools promise great outcomes – Vapi has delivered on them,” said Jason Mitura, vice president of software development at Amazon Ring. He noted that Ring’s customer satisfaction scores improved after deploying Vapi’s platform and that teams could tune the AI agent experience without depending on engineering resources.
The startup’s origin story is unconventional. Dearsley built an AI therapist in 2023 for conversations during his daily walks. While few people wanted the therapy product itself, startups kept asking about the low-latency voice infrastructure powering it. This led the pair to pivot from their Y Combinator productivity startup Superpowered to launch Vapi publicly in 2024.
Vapi now handles more than 1 billion calls through its platform, processing between 1 million and 5 million calls daily. The bulk of that volume comes from enterprise customers including:
- Amazon Ring
- Kavak
- Instawork
- New York Life
- Intuit
- Cherry
The company also operates a self-serve developer platform used by more than 1 million developers. “Because we started from self-serve and had such a wide developer footprint, we were already battle-tested at significant scale before we signed our first major enterprise customer,” Dearsley said.
Peak XV Partners led Vapi’s Series B round, with participation from Microsoft’s M12, Kleiner Perkins, and Bessemer Venture Partners. The funding brings Vapi’s total raised to $72 million, with the startup currently at an annual recurring revenue run rate in the “healthy” eight figures, according to investor sources.
Vapi faces competition from a growing wave of AI voice startups including Sierra, Decagon, PolyAI, Bland, Retell, and ElevenLabs. The market is heating up as companies race to build systems that can handle customer conversations with minimal human involvement. Dearsley said Vapi stands out by focusing on infrastructure and orchestration rather than pre-built applications, particularly for enterprises wanting control over reliability, compliance, and model behavior.
The timing appears right for AI voice adoption. Businesses are under pressure to reduce customer service costs while maintaining quality, and recent advances in large language models have made voice agents more reliable and natural-sounding. Ring’s success with Vapi could encourage other large enterprises to make similar moves.
With around 100 employees, Vapi plans to use the new funding to expand its engineering, infrastructure, and go-to-market teams. The company is betting that more businesses will follow Ring’s lead in moving customer interactions to AI systems.
“The golden problem is taking this indeterminate beast that is a model and taming it,” Dearsley said. “If you can do that, then you can provide value to the world.”
