PolyAI Raises $50M Series C to Bring Human-Like Voice AI to Enterprise Customer Service
April 22, 2024
PolyAI, the London-based enterprise voice AI company, has raised $50 million in a Series C funding round. The round is backed by new investors Hedosophia, NVIDIA's NVentures, and Zendesk, alongside follow-on investment from existing backers Khosla Ventures, Georgian, Point72 Ventures, Sands Capital, and Passion Capital. The raise brings PolyAI's total funding to over $120 million.
Founded in 2017 as a spin-out from the Machine Intelligence Lab at the University of Cambridge, PolyAI was built by CEO Nikola Mrkšić, CTO Tsung-Hsien (Shawn) Wen, and Pei-Hao Su — all of whom completed PhDs in machine learning and dialogue systems. The co-founders had front-row seats to the development of the first transformer-based models, and chose to apply that expertise not to consumer products but to enterprise contact centres, where customer service automation had long underdelivered on its promise.
Unlike most conversational AI firms that bolted voice onto text-based chatbots, PolyAI built its own speech recognition engine and large language models tuned specifically for the controlled, brand-consistent performance enterprises demand. The result is a voice assistant capable of understanding callers regardless of accent, pace, or vocabulary — and handling entire service interactions without escalation. Customers including FedEx, Marriott, Caesars Entertainment, and Unicredit have deployed PolyAI across millions of calls.
The company has grown 10x over the past two years and generates over $30 million in annual revenue, with more than 70% coming from the US. NVIDIA's NVentures participation reflects the chipmaker's growing interest in enterprise AI infrastructure — PolyAI was one of a select group of European companies Nvidia backed in 2024 and 2025.
CEO Mrkšić has said the company aims to handle half of all automated customer service calls globally within five years. The Series C capital will be used to advance PolyAI's proprietary AI models, expand go-to-market in the US and Europe, and continue building out its Agent Studio platform.
Sources





