Skip to page content

Cambridge 'Voice Skins' Startup Modulate Raises $2M in Seed Funding


Modulate_1
Image credit: From left, Modulate CEO and co-founder Mike Pappas, VP of audio Terry Chen and CTO and co-founder Carter Huffman. (Photo courtesy of Modulate)

Modulate, a Cambridge-based startup producing voice features for digital characters in video games and other online experiences, raised a $2 million seed financing round co-led by 2Enable Partners and Hyperplane Venture Capital.

The software startup, which was self-funded since its inception in August 2017, uses machine learning techniques to offer video games players an array of customization features for the voice of their avatars. It was founded by two MIT grads, Mike Pappas (an early employee of Paul English’s travel startup Lola) and Carter Huffman, who recently hired Terry Chen as VP of audio.

When researching investors, Modulate got in touch with Boston-based angel Michael Dornbrook. He's the former COO of Cambridge video game development company Harmonix – the company behind Rock Band and Guitar Hero games – and a board member of video game studio Disruptor Beam.

"We first sat down with him as an individual angel," Pappas explained. "He's worked extensively with 2Enable... on the Harmonix investment. He reached out to John [Huysmans], who's the managing partner there... By the end of that lunch, we were sort of coming out of it as, 'Yes, we all definitely want to do this deal together.'"

Many video games, including vastly popular Epic Games' Fortnite, include features—sometimes available as in-game purchases—that allow players to change hair, skin color, clothes and accessories of their virtual personae. Modulate offers software development kits that can potentially be integrated into any platform, with the final goal of giving players' voices a makeover while maintaining things like cadence and inflections. If you're a man, you can sound like a woman, and the other way around.

Such "voice skins," as the company defines its product, can also have malicious applications. The more machine learning and artificial intelligence software become sophisticated, the harder it gets to distinguish between authentic content and "deep fakes." Technology can already make it look like anybody - think Donald Trump, Barack Obama, Hillary Clinton - has said or done anything, generating misinformation. That's why Modulate always includes a watermark on the voice clips they produce.

“Modulate is about creativity and freedom, not impersonating others," Huffman said in a statement. "We've built ethical safeguards into our company from the ground up, from how we distribute our technology, to how we select the voice skins to offer, to watermarking our audio for detection in sensitive systems."

Modulate plans to use the funds to expand its team, starting with a couple more engineers with experience in machine learning, C++ and deep neural networks.


Keep Digging

Allium SJ, SM Mill photo edit
Fundings
Ivan Cheung
Fundings
Rahul Kakkar, Tome Biosciences
Fundings
Leah Ellis Yet Ming Chiang photo
Fundings
Nick Harris
Fundings


SpotlightMore

See More
See More
See More
See More

Upcoming Events More

Jun
14
TBJ

Want to stay ahead of who & what is next? Sent daily, the Beat is your definitive look at Boston’s innovation economy, offering news, analysis & more on the people, companies & ideas driving your city forward. Follow the Beat.

Sign Up