voiceitt

Voiceitt develops automatic speech recognition technology designed to understand non-standard speech — speech affected by conditions such as cerebral palsy, stroke, Down syndrome, head and neck cancer, Parkinson's disease, ALS, multiple sclerosis, muscular dystrophy and Huntington's disease, as well as accented speech, aging voices, and speech of Deaf and hard-of-hearing people who use spoken English. The core product is a browser-based application that runs on any internet-connected device with a web browser, combining speech-to-speech (repeating the user's utterance in a synthesized voice) and speech-to-text dictation intended to work on spontaneous speech rather than only pre-trained phrases.

The product line includes Voiceitt for Chrome, a speech-to-text input delivered through a Chrome extension; real-time captioning add-ons for video meeting platforms — Webex (via a partnership with Webex by Cisco), Zoom, and Microsoft Teams, where meeting hosts or participants add the "Voiceitt Captions" app to a meeting or chat and sign in with a Voiceitt account, with guests able to view another user's captions; and integrations with smart assistants including Amazon Alexa, Google and Siri, plus AI productivity tools. Voiceitt also offers a Speech API and SDK, promoted with a developer portal, aimed at making IVR phone systems, voice assistants and other voice interfaces usable by people with dysarthria.

Users personalize the system by recording phrase cards — a minimum of 50 recordings unlocks the generic recognizer, with continued recording improving accuracy — and can add personalized vocabulary cards for proper nouns or jargon and shortcuts that expand a short spoken trigger into a longer spoken message. The company states the technology is most effective for mild to moderate speech impairments and can offer up to three times better recognition rates than standard speech recognition software. Recordings are de-identified and linked to an anonymous ID, with only the customer success team able to link recordings to a user's email and personally identifiable information; deleting an account deletes the user's PII from the database.

Founding story

Co-founder Danny Weissberg's grandmother had a stroke and could no longer be understood by her family, though a nurse could understand her. That observation led him to conclude that AI could be trained to interpret non-standard speech, and the company was founded and developed in Ramat Gan, Israel. Co-founders Danny Weissberg and Stas Tiomkin are both Technion alumni.

Business model

Voiceitt sells subscription access to its application and meeting-captioning add-ons to individual users and organizations, with a free 30-day trial requested through support, and licenses its speech recognition to developers and manufacturers through an API and SDK for integration into voice products and IVR systems. Meeting integrations require a Voiceitt account and, for Webex, a paid add-on that may entail additional cost; the Teams captions app requires a paid Microsoft 365 account.

Paid Voiceitt accounts and add-ons for meeting platform captioning, plus API/SDK licensing to partners and manufacturers integrating the speech recognition into their own products.

Traction

Reported integrations and partnerships with Amazon Alexa, Webex by Cisco, Microsoft Teams and Zoom; an iOS app with Alexa integration; deployments with the State of Tennessee Department of Intellectual and Developmental Disabilities under its DIDD Waiver programs and with disability service provider St. John's Community Services in Washington, DC; participation in the European Commission "Nuvoic" project with the Karten Network in the UK; and a beta program for its spontaneous-speech technology drawn from its network of nonprofits, hospitals, clinics, government entities and educational institutions.

Latest developments

Product line expanded to meeting-platform captioning — Webex through a partnership with Webex by Cisco, plus Zoom and Microsoft Teams (the Teams captions app is listed on the company site as coming soon) — alongside Voiceitt for Chrome and a Speech API with a developer portal aimed at making IVR systems accessible to people with non-standard speech.