How a schoolgirl from Almaty created a startup that helps overcome speech barriers
Ukili Omarova is a schoolgirl from Almaty who, at the age of 15, founded the startup Soile AI. Today, her team of four girls is developing technologies capable of recognizing impaired speech and helping people around the world speak confidently again.
Ukili Omarova, city — Almaty, Founder and CEO of Soile AI, Instagram
How it all started
I’ve been working on projects since I was 12 — back then they were mostly scientific research. Over time, I wanted to show my hypotheses through code, and that’s how I gradually came into the IT field. In August last year, I learned about startups and began creating my own.
When I was little, my grandfather had a stroke, and after that, he lost his speech abilities. We all stopped understanding him. He faced issues like social isolation and loss of self-worth. Ten years passed. Then I joined a hackathon where we had to develop inclusive technology — that is when I remembered his story, and an idea came to me. I did a bit of research and found out that more than 800 million people around the world suffer from speech impairments. In CIS countries — more than 50 million, and this number continues to grow, especially after the pandemic. People with such problems often face social isolation, loss of self-esteem, and many other difficulties. These issues are caused by more than 300 conditions — from cerebral palsy and Parkinson’s to post-stroke states.
Imagine being surrounded by your closest people but still feeling alone because no one can “hear” or understand you. That is a serious problem.
Now imagine there is a smart assistant that understands you. That is Soile AI — artificial intelligence capable of recognizing impaired speech and effectively helping in its recovery.
Let me show you a real example. We have a Telegram bot that functions as a speech recognizer, and a web platform with extended features: recognition, speech recovery, and an AI advisor.
How it works? First, you choose a language — we support a multilingual interface and are focused on CIS languages. Then, you record a voice message. Our bot analyzes it and within seconds provides a text transcript and an adapted audio version in clear speech.
In addition, we integrate with smartwatches for faster and more convenient access to features.
But we don’t stop there. On our web platform, you can also use the speech recovery feature. Additionally, we offer:
— a library of video workouts: both long and detailed, and short in story format
— live consultations with specialists — speech therapists, psychologists, and others
— an AI advisor you can contact at any time with speech-related questions
We combine professional experience with the capabilities of artificial intelligence to provide every user with accessible, personalized, and most importantly — high-quality speech recovery support.
Startup launch
Exactly six months ago, I only had an idea and a minimal product prototype. I was the only person in the team and just wanted to share the concept — to see how valuable it might be. That is how I won a hackathon on inclusive technologies using AI. IT-FEST 2024 — the largest festival in Kazakhstan, where participants from all over the country come to present their projects. I was the only schoolgirl there — and I took first place in the startup battle.
Later I won first place in the CIS at a social startup competition — again as the only schoolgirl among the participants.
Then I realized: to truly implement the idea, I needed a strong team. I started gathering other school students from public schools. The path was not easy — there were many challenges. I learned how to manage people and continue learning — it is a separate, complex science. I mastered how to choose the right teammates, control processes, delegate tasks, maintain team spirit — and much more.
In our team, there are four girls. I am the founder and CEO of the startup, I oversee all directions and work in multitasking mode. We have a programmer, a team member knowledgeable in medicine, and a specialist in social media promotion.
Target audience
We have a wide and flexible target audience. These are not only people with speech disorders, their loved ones and specialists.
Our platform is also useful for:
— those learning a foreign language and struggling with an accent — the bot will show where you mispronounce and how to sound better
— those who want to speak confidently in public — since many stutter, stumble, or lose confidence on stage. We solve this problem with public speaking video courses available on the platform.
In the B2B segment, we collaborate with private medical centers, schools for children with special educational needs, and public speaking academies.
Currently, 14 private medical centers and a public speaking school are working with us. BI Group has also shown interest in integrating our solutions.
The B2G segment includes public clinics, educational institutions, and community spaces.
This summer we are conducting pilot tests to expand the voice database that trains our AI.
Incubation programs
We have not gone through any incubators or accelerators. We do not even have a mentor — we do everything ourselves, learning through video lessons and online materials.
As of today, we have several investors, but our priority is government grants, as we consider them a more beneficial and stable funding source at the early stage.
Challenges
There have been many difficulties on our path that we do not often speak about. For example, we were not always taken seriously because of our age. Just imagine: not long ago I was 15, coming to medical centers with partnership proposals. Age is both an advantage and a challenge.
The second challenge — we are girls. In our society, there’s still the belief that tech is not “a women’s thing”. But we’re four girls in the team. When we hear such stereotypes, we just smile and keep doing our work. Hopefully, in the future, such beliefs will disappear.
My advice to beginners — never give up. Before Soile AI, I had four startups in six months. I kept losing, losing teams — many stopped believing we’d ever win or launch a project.
But I kept going. So — try again and again. Do not stop. Go all the way, especially if you feel this is truly yours. Do what brings you joy — that is what gives you strength and moves you forward. And of course — be in the right place at the right time.
Achievements
We are the top‑1 startup in Kazakhstan, the CIS, and Asia in our category. Our project caught the attention of the International Artificial Intelligence Association, and we are invited to speak at global forums.
We have already won more than 10 international startup battles and participate in global expos alongside major tech companies.
And most importantly — everywhere, we are the only school students, and I am truly proud of that.
Countries like Kyrgyzstan, China, Korea, Russia, the USA, Uzbekistan, and Japan have already shown interest in our project.
Plans
This summer we’re conducting pilot tests and continuing to expand our voice database. We aim to present our project on different platforms to help as many people as possible.
Our plans for the next year — entering neighboring markets, starting with Uzbekistan and Kyrgyzstan.
At one of the international forums, deputy ministers mentioned our project — which was a pleasant surprise for us. Many respected figures from Kazakhstan were amazed by our work and expressed readiness to support us. Additionally, American researchers from top universities are also interested and want to support our project.