How a Ukrainian IT engineer uses Google AI to help children with speech disabilities find their voice

An interview with Sergij Rylskyj, co-founder and CTO of Enhance Your Ability or EYA, a Ukrainian startup developing an inclusive communication platform for children with disabilities.

When my daughter was born with cerebral palsy, we held on tightly to the hope that she would be able to communicate with us one day. We visited many doctors and psychologists all over Ukraine, but found no easy answers. At first, we were hesitant to try alternative communication methods — we just wanted to hear our daughter speak. Eventually, we realised what’s more important is giving her the tools to express herself, in the way that she needs them. We tried various apps, but they were neither effective nor affordable. That’s when I turned to my wife and said, “We can do better.” We made this our mission: to give children with speech disabilities, like my daughter, the freedom to communicate easefully.

With 13 years of experience as an IT engineer, I began exploring AI in my spare time, driven by curiosity. Then I attended the Machine Learning Bootcamp run by Google in Ukraine. It made me realise, wow, I can create such a system, from home in my own time.

I developed EYA, a picture-based communication platform to empower children with speech disabilities. Children simply select digital picture cards to build their thoughts, and the AI instantly transforms those selections into spoken sentences. It’s a moment of pure magic: Suddenly, they have the power to be heard. My wife and I decided that the core app should be free for everyone. You simply download it to your device and your child can begin expressing themselves. They can use the app to say, “I want water” or “I love you.” The app also suggests cards, depending on where the child is — school, home, or with their therapist.

Behind the scenes, we rely on Google’s infrastructure to power this experience. We leverage Google Cloud Text-to-Speech to give children a voice, Google Cloud Speech-to-Text to transcribe therapy sessions, Gemini Enterprise Agent Platform so that app users can search for what they’re looking for with keywords, Gemini for content generation, and Firebase to handle all the backend heavy lifting like authentication, storage, and hosting.

“With Google AI and the cloud credits we received through the Google for Startups Cloud Program, we've been able to focus on our idea without cost being a barrier.”

We’ve now expanded the platform to create a more connected ecosystem which links everyone around the child — from family members to therapists, clinics, and the school. For example, a therapist can use Google Cloud Speech-to-Text to leave notes in the app for a parent to update them about their child’s session. This makes it effortless for parents to stay informed, providing a clear, real-time window into their child's progress. Additionally, we added a feature where a parent or teacher can take a photo of a textbook page and AI adapts it into accessible content. We’ve already reached 2,000 users and hope to reach many more as 43 percent of children with clinical communication disabilities don’t have access to additional tools of communication like these. Usually, such a platform would be impossible for a small team to create, but we’re able to do it because AI helps review the code, build pipelines and generate content.

“Leveraging Google’s ready-made AI empowers our small team to punch above its weight.”

Any new features we add are inspired by my daughter’s needs. In the end, everyone benefits from inclusive communication. Today I can communicate with you, but tomorrow, that could be a different story. One reason is the war in Ukraine — but, it’s a reality any one of us could face. Inclusive speech empowers everyone — stroke survivors, people recovering from accidents, or anyone facing communication challenges.

Looking ahead, some children, like my daughter, need eye-tracking systems to interact as she can’t use her hands. At the moment, this costs €15,000, which many families can’t afford. In the future, I hope to develop affordable eye-tracking systems. Google AI will help provide the data to train models and storage. We’re also using Gemini to adapt educational content, in a way that’ll allow children with special needs to study more effectively. We’re committed to empowering children everywhere, working toward a world where no child is left without a voice.