Voice-controlled technology like Amazon Echo, Siri or hands-free features in Google Maps are things we’re starting to take for granted. But as Mary Meeker’s 2017 Internet Trends Report noted, voice controls are changing computer-human interfaces, and industries, broadly. Speech recognition or voice controls are being added to medical devices and business applications, even vehicles and industrial robotics.
But there’s a problem — voice systems have been built for standard speech today. That leaves out millions of people who live with speech impairments, or who just have a strong accent. Now, a Tel Aviv-based startup called Voiceitt has raised $2 million in seed funding to translate into clear words speech that’s not easily intelligible.
The startup, which was co-founded by CEO Danny Weissberg and CTO Stas Tiomkin, is a graduate of the DreamIt Health accelerator. Investors in Voiceitt’s seed round include Amit Technion, Dreamit Ventures, Quake Capital, Buffalo Angels, 1,000 Angels and other angels.
Here’s how Voiceitt works: Users fire up the company’s app and it asks them to compose then read short, useful sentences out loud, like “I’m thirsty,” or “Turn off the lights.” The software records and begins to learn the speaker’s particular pronunciation. A caregiver can type phrases into the app if the user is not able to do so independently.
After a brief training period, the Voiceitt app can turn the user’s statements into normalized speech, which it outputs in the form of audio or text messages, instantly. Voice-controlled apps and devices can easily understand the newly generated audio or written messages. But Voiceitt also can be used to help people with speech impediments communicate face to face with other people.
Dreamit’s Karen Griffith Gryga said investors view Voiceitt as a technology that’s starting with “the thin edge of the wedge,” in the market for assistive tech. But it could be expanded to help people with strong accents use whatever voice-enabled technology Seattle or Silicon Valley comes up with next.
Weissberg explained that he came up with the idea for Voiceitt after his grandmother suffered from speech impairments following a stroke. The CEO said, “I realized how we take for granted the way we communicate by speaking. Losing this is really terrible, one of the hardest aspects of stroke recovery. So I didn’t say, right away, let’s start a company. But I began to talk with speech therapists and occupational therapists, and to learn everything I can about the problem and whether there was a market in need, there.”
An early version of Voiceitt will be available next year, but the app is in beta tests now. The company’s pilot customers are hospitals and schools, and people there who have speech differences because of a health condition, like those with cerebral palsy, Down syndrome, Parkinson’s or who are recovering from a traumatic brain injury or stroke.
Long-term, Weissberg said, “This could really be an accessibility extension to speech recognition for anyone, Google, Amazon, Apple, IBM or Microsoft. We’d love to function like a major OEM and work with all the major platforms.”