PlayAI Lands $21M Seed to Scale Its Human-Like AI Voice Models
AI voice startup PlayAI announced it has raised $21 million in a Seed round. The company says it will use the capital to invest in its generative AI voice models and voice agent platform.
“Speech as an interface is exploding in popularity, and we knew it was a massive opportunity from the get-go,” said Mahmoud Felfel, co-founder and CEO of PlayAI, in a release. “Building voice agents that can converse like humans and autonomously handle complex tasks is no easy feat, and I'm immensely proud of what our team has achieved. This funding will help us deliver our vision of powerful, emotive, and human-like voice interfaces for any application.”
PlayAI (formerly known as PlayHT) enables developers to build voice applications using the company’s custom LLMs which it says are trained on an extensive dataset of diverse human speech containing podcasts, narrations, storytelling, and business conversations. Users can clone voices across multiple languages and accents using the company’s AI models which are accessible through text-to-speech APIs.
AI agents, or autonomous software designed to perform tasks on behalf of users, are gaining significant attention for their versatility across industries. Responding to this demand, PlayAI offers Play 3.0 Mini, a smaller model specifically designed for training AI agents. Supporting more than 30 languages and boasting specialized capabilities in handling acronyms and numeric sequences, Play 3.0 Mini enables users to quickly create generative AI voice agents tailored for applications such as customer support, appointment scheduling, and sales lead engagement.
The front page of PlayAI’s website features a demo of its voice agents where users can test out conversations with an AI healthcare appointment scheduler, an e-commerce store, a hotel concierge, and even a neuroscientist with an icon that looks suspiciously like Andrew Huberman.
The company believes the generative AI voice industry will grow fourfold in the next decade, citing a Market.Us report in its announcement. The company says previous generations of voice technology gave unnatural results due to speech factors like incorrect pacing, emphasis and cadence.
(Source: PlayAI)
Along with its funding announcement, PlayAI also unveiled a new version of its speech model PlayDialog, a multi-turn text-to-speech model the company says uses a conversation’s historical context to control rhythm, intonation, emotion and pacing to deliver more natural sounding speech. PlayAI claims PlayDialog was trained on hundreds of millions of conversations that represent real-world examples and can deliver human-like conversation with natural delivery and appropriate tone in real time. This model can be used via API or through a new tool called PlayNote that transforms PDFs, text, or videos into stories, podcasts, or briefings.
Today’s $21 million Seed round was led by Kindred Ventures and 500 Global with participation from Race Capital, Y Combinator, Soma Capital, Pioneer Fund, TRAC, and others. As part of the announcement, Steve Jang of Kindred Ventures joined as a board observer.
Race Capital General Partner Chris McCann commented in a release that the AI voice market has the potential to become a $2 trillion industry that could have a transformative impact across many sectors.
“Play AI’s voice AI platform is the key to unlocking new applications across customer support, sales, marketing, and beyond. We couldn’t be more excited to partner with Mahmoud, Hammad, and the PlayAI team on this journey.”
“We’ve been early big believers in the nascent and rapidly-evolving generative media space,” said Steve Jang, founder and managing partner at Kindred Ventures. “AI voice generation platforms are fundamentally transforming how enterprise and consumer businesses are communicating with their customers, and we're proud to back PlayAI to further the development of their powerful mission.”
