Amazon Ring has shifted its customer support operations toward AI-driven voice systems, partnering with startup Vapi after testing more than 40 vendors to handle rising call volumes during the holiday season.
The rollout now routes all inbound customer calls through Vapi’s platform, replacing traditional call-center and automated phone systems with AI agents designed for more natural, responsive conversations. The move was driven by Ring’s need to scale support efficiently while maintaining service quality during peak demand.
Following the deployment, Vapi raised a $50 million Series B funding round led by Peak XV Partners, valuing the company at around $500 million post-investment. Other investors in the round include Microsoft’s M12, Kleiner Perkins and Bessemer Venture Partners, bringing total funding to about $72 million.
Vapi was co-founded by Jordan Dearsley and Nikhil Gupta after experimenting with an AI-based therapy product in 2023. While the original idea did not gain traction, it led them to focus on building low-latency voice infrastructure for startups and enterprises. The platform officially launched in 2024.
Today, Vapi provides tools for building and managing AI voice agents used in customer support, sales, scheduling and lead qualification. The company says it processes between 1 million and 5 million calls daily and has handled over 1 billion calls overall, with enterprise clients driving most of the usage.
Alongside Amazon Ring, Vapi’s customers include companies such as Kavak, Instawork, New York Life, UnityAI, Cherry and Intuit. It also supports a developer platform used by over a million developers, helping it scale before signing major enterprise deals.
The company currently has around 100 employees and plans to expand its engineering and go-to-market teams using the new funding. Its broader position is in the growing AI voice sector, where multiple startups are competing to build systems capable of handling human-like customer interactions at scale.
Vapi’s focus, however, is on infrastructure rather than packaged applications, giving enterprises more control over reliability, compliance and behavior of AI agents.
