Using artificial intelligence, Voiceitt — an AgeTech Collaborative™ (ATC) startup participant has built proprietary automatic speech recognition technology to help people with speech disabilities, disorders or impairments. Its first product, a mobile app, enables those with nonstandard speech to access voice-activated devices such as Siri, and to communicate by voice and be understood. Voiceitt also offers a unique integration with Amazon Alexa, enabling those with speech and motor disabilities to access Alexa independently.
We spent time with Sara Smolley, co-founder and vice president of partnerships and business development, to learn more about the company, its breakthrough technology and exciting recent developments including something new with Webex.
This interview, originally a phone interview, has been edited for clarity and length.
Please tell us about Voiceitt.
We’re a mission-driven artificial intelligence (AI) company, and our core technology has the ability to recognize nonstandard speech; clinically, this is called dysarthric speech. Our mission is to leverage our accessible and inclusive voice AI to help aging adults and people with disabilities who have nonstandard speech to live more connected and independent lives.
Dysarthric speech is often correlated with underlying medical conditions, disorders and disabilities, including cerebral palsy, different forms of autism, Parkinson’s and multiple sclerosis, as well as a variety of other conditions, such as stroke and brain injury.
How do consumers access Voiceitt?
Our product is available in two ways. The first is our API, which is the capability to be integrated into any voice interface, such as smart speakers and voice-activated devices. This can enable a person with nonstandard speech to be understood in a conversation or video conference, or enable them to interact with other voice technologies. For example, if a person wants to turn on the lights using a voice assistant, they can use Voiceitt to “translate” their command to the voice assistant to carry out the command.
Our second product is our consumer application. This is a browser-based application that anyone can sign up for online, and it can be used on any device with a browser. The user trains the software with their unique pronunciation, and then they can dictate notes, emails and social media posts, or use it for in-person communication.
Can you tell us about your partnership with RAZ Mobility, which is also an ATC startup participant?
RAZ Mobility is going to be the exclusive reseller of our voice application in North America, selling the app directly to individuals and institutions such as special needs schools, assisted living facilities and rehabilitation hospitals.
We’re also planning a second, parallel partnership with RAZ Mobility where our technology is integrated into their phone. That will enable people with nonstandard speech, including aging adults, to use the phone and to have their voice transcribed during phone calls.
We are excited to witness the creative possibilities our community members will explore, enhancing productivity, fostering connections with loved ones, and bringing their ideas to life using their voices. We couldn't have found a better partner than RAZ Mobility to put Voiceitt 2 in the hands of users.
You’ve also recently announced a partnership with Webex.
Cisco, which owns Webex, has a mission to help create more inclusive workplaces and promote inclusive collaboration. The first step in our partnership with them is to make Webex accessible by enabling employees with disabilities to participate in work meetings—especially hybrid work environments, where having transcription over a video meeting is essential for them to participate at work. We’re also working on a native integration, which will be a deeper, more seamless integration with the Webex product.
What has stood out to you about your collaboration with innovators such as RAZ Mobility and corporate organizations such as Webex by Cisco?
As part of our company’s journey, I’ve learned that the unique nature of this technology means that success hinges on the shared learning of innovators in government, nonprofit and corporate partnerships. Our partnerships are a snapshot of how innovation in this field requires input from all these stakeholders.
We’re proud that we’re not just building another cool technology, but that we’re addressing the needs and preferences of the people that we support and serve. We couldn’t do that without the support of the network of relationships we’ve made, such as with AARP, our pilot partners, our corporate partners, and others.
What’s your “secret sauce” for all this success?
One ingredient is the huge database that we’ve been building. It has voice recordings of people with dysarthric speech that have been processed, annotated and analyzed through our unique transcription methodologies. The database continues to grow: We have a consent process that allows us to anonymize peoples' voice recordings and use them to augment our algorithms to enable the technology to work better.
Another ingredient in our secret sauce is the incredible knowledge and expertise we’ve built over the course of a decade learning about the unique idiosyncrasies of people with dysarthric speech. This includes our team. We employ people with nonstandard speech as early adopters and as core team members. Including our employees with disabilities — and stakeholders in the disability community, more generally — in our process allows us to accelerate the development of the technology.
What was the original inspiration behind the company?
Voiceitt resonates on a personal level for our founding team and our core team. In my case, it was my grandmother, who was diagnosed with early-onset Parkinson’s disease. A very diverse set of conditions affect speech and disability, and it really comes back to personal experiences.
But as we reach these significant milestones building a complex technology in a challenging industry, it’s not just our personal stories that inspire us, but the individuals who have become part of our communities, and the hundreds of people who have reached out to us and told us how meaningful Voiceitt is to them and their families.
Is there anything else you’d like to mention?
It used to be that if a person with nonstandard speech was training the machine learning algorithm on their unique way of speaking, it would take a thousand voice recordings, which takes quite a bit of time for the user. But we’ve launched a next-generation technology that dramatically reduces the amount of time that training process takes, while still being personalized to the user. This gets us closer to our goal of opening up mainstream products to people with disabilities.
On that note, I’d like to highlight that there’s a piece of new legislation being brought before Congress. It’s called the Communications, Video, and Technology Accessibility Act of 2023. It’s a landmark piece of legislation that essentially revamps the Americans with Disabilities Act: Where the concern used to be getting a ramp for an apartment building, this legislation is like getting a ramp to access a website. We’re proud that Voiceitt has been involved with advancing the technology for this, and together with policymakers, it should really benefit the entire disability community.