PolyAI Raises $86M at $750M Valuation to Scale Conversational AI for Enterprise Call Centers

Call center chatbot startup PolyAI raises $86M at a $750M valuation - SiliconANGLE

London-based artificial intelligence startup PolyAI Ltd. said today it has raised $86 million in funding at a $750 million valuation for its agentic chatbot platform that automates customer interactions for casinos, banks, hospitals and other enterprises. Today's Series D round was led by Georgian, Hedosophia and Khosla Ventures, and saw participation from Nvidia Corp.'s venture capital arm NVentures, plus the British Business Bank, Citi Ventures, Squarepoint Ventures, Sands Capital, Zendesk Ventures and Point72 Ventures. The round brings its total amount raised so far to more than $200 million. The company was spun out of the University of Cambridge after developing AI assistants for call centers that can guide human customers through various enquiries and handle millions of calls daily. Moreover, its call center agents are said to be almost indistinguishable from human voices. PolyAI co-founder and Chief Executive Nikola Mrksic said that the company's vision has always been to make AI voices sound so ultra-realistic that every enterprise wants to use them. "PolyAI started with a simple idea: enterprises should sound human," he explained. "We turned that idea into reality, and it led to something far greater: the emergence of the agentic enterprise. This is a living, breathing system that understands what your customers, employees, and AI agents are doing in real time." The startup said it is now handling customer calls about power outages for California's largest energy firm, Pacific Gas and Electric Co., as well as inquiries about new and lost credit cards for the Italian bank UniCredit S.p.A. and restaurant bookings for Las Vegas casinos such as Caesars and Golden Nugget. Mrksic told Forbes in an interview that voice AI isn't just about reducing costs by replacing human agents in call centers, but helping enterprises to drive revenue by automating tasks that aren't economical for humans to handle. "Our AI picks up every phone call, they always book you in and they never forget to upsell you," he said. "It helps companies become the best version of themselves." Mrksic, who previously worked on Apple Inc.'s Siri voice assistant, said the realism of PolyAI's voices and their flexible nature has enabled PolyAI to grow its sales significantly over the last year. He also told Forbes that the company is on track to double its annualized recurring revenue in the financial year to more than $40 million, with sales to U.S.-based clients expected to triple. Because PolyAI is based in the U.K., it's legally required to publish financial records, and these show that its revenue grew from $8.9 million in the 12-month period ending January 31, 2024 to just over $15 million by the end of the same period this year. PolyAI is rivalled by a number of U.S. voice AI startups, and they have also attracted massive amounts of venture capital in recent months. For instance, Sierra Technologies Inc. closed on a bumper $350 million funding round in September that increased its valuation to over $10 billion, while San Francisco-based Decagon AI Inc. got $131 million in a June round that valued it at $1.5 billion. It faces competition in Europe too, with Paris-based Parloa GmbH raising $120 million at a $1 billion valuation in May. One of the main differences between PolyAI and its rivals is that it develops its own large language models, which is what gives it an edge in terms of realism, Mrksic said. Its competitors simply plug into LLMs from companies such as OpenAI Group PBC or voice-specific AI startups such as Eleven Labs Inc. PolyAI's $750 million valuation is based on a 25-times revenue multiple, which is relatively conservative compared to some of its U.S. competitors, whose valuations are based on 100-times multiples. But in any case, each of them is way north of traditional listed software companies, where valuations tend to be based on multiples of around eight-times revenue. The VCs backing PolyAI are clear about the generational opportunity of AI, however. Emily Walsh, lead investor at Georgian, said this is the second time she has returned as an investor in the startup. "For the world's largest brands, customer service is no longer just a cost center, it's a massive opportunity for value creation," she said. "PolyAI's ability to deploy lifelike voice agents at enterprise scale unlocks significant savings and revenue." Going forward, PolyAI plans to spend money on further developing the proprietary technology that powers its Agent Studio platform and expand its go-to-market teams.

PolyAI Raises $86 Million to Bolster Conversational AI Platform | PYMNTS.com

"This is a living, breathing system that understands what your customers, employees and AI agents are doing in real time -- and helps them all succeed together," PolyAI co-founder and CEO Nikola Mrkšić said of the company's technology in the release. "Like seeing a single drop of water and fixing a leak before it bursts, the agentic enterprise can detect and respond to problems and opportunities before human agents even know they exist. We're building this future with the world's leading enterprises, where our AI helps millions of customers every day." "For the world's largest brands, customer service is no longer just a cost center, it's a massive opportunity for value creation," Emily Walsh, lead investor at Georgian, said in the release. "PolyAI's ability to deploy lifelike voice agents at enterprise scale unlocks significant savings and revenue." The funding round came as retailers treat conversational AI as a "new commerce surface" where "content, discovery and payments converge," PYMNTS wrote last week. PYMNTS Intelligence found that 42% of shoppers used AI assistants during Black Friday this year to find discounts, 35% to monitor prices and 31% to compare products. Major retailers have begun integrating with OpenAI to allow for shopping inside ChatGPT. Smaller merchants are restructuring product listings to show up in AI-generated answers instead of traditional search results. Still, there are limits to communicating with AI via voice, Mrkšić said in an interview with PYMNTS last year, as AI struggles to recreate human empathy and emotional intelligence, which can make interactions feel cold and impersonal. "If someone crying calls an AI-powered customer service line, the AI will treat them exactly the same as any other caller because that's what it's programmed to do," he said. "Additionally, as with all technology, there are security risks associated with unsecured voice AI. Those implementing voice AI must be wholly cognizant of the technology's limitations and recognize the likely need for appropriate safeguards."