Israeli speech recognition startup, Voiceitt, released their latest product, Voiceitt2, a new application geared for people with speech disabilities. Partnering with assistive tech company RAZ Mobility, the application translates non-standard speech into standard speech, aids in transcription, dictation, and connects with AI assistants like ChatGPT. An additional feature includes Webex integration for video session transcription.

Voiceitt2 will be available initially in North America in English as a browser-based software. Extensions to the UK, Australia, and other languages are slated for 2024.

RAZ Mobility, known for its assistive mobile technology, will handle the distribution. Both firms are part of AARP’s AgeTech Collaborative.

Voiceitt offers a unique and inclusive voice-enabled experience for individuals with non-standard speech patterns. Through the Voiceitt API and the Voiceitt2 web-based application, users are guided by simple on-screen prompts to train the software, ensuring it can accurately recognize and transcribe their unique speech patterns.

The startup’s technology achieves high accuracy with little training data, rivaling the experiences provided by leading global speech recognition systems for standard speech. It extends speech-to-text functionalities to those with dysarthric or atypical speech, including applications like meeting captions, voice assistants, and smart home controls. Moreover, Voiceitt supports swift, accessible training interfaces through algorithms specifically designed for high accuracy even with scarce data. As users engage with the system, its machine learning mechanisms continually refine its accuracy, minimizing the required training duration.

Voiceitt, founded by Danny Weissberg in 2012, has raised a total of $20.5 million in funding. Their latest round, Series B, which closed last year, brought in $4.7 million from investors including Third Culture Capital, AMIT Technion and Cisco Investments.