Our Journey
Before ChatGPT, we built one of the first conversational toys for kids. We set up in the U.S. and tested with real families.
We spent time in daycare centers, kindergartens, and living rooms. We watched kids play. We listened to how they talk. We saw how fast trust breaks.
The same problems kept showing up:
- Speech recognition failed on young voices
- Safety was an afterthought
- Privacy couldn't be patched in later
- Characters acted unpredictably
- No one could explain the system to kids or parents
The infrastructure didn't exist.
So we started building it. From scratch. With kids involved the whole way.
That work became the UG Kids Lab - ongoing testing with real children, across products and settings. What started as qualitative research grew into data collection at scale. Hundreds of thousands of consent-based child interactions.
That data now powers the first child-tuned speech and language models in the world.
What We Believe
Kids deserve a playground. Open, inviting, safe. A place to explore freely, within clear boundaries.
Parents belong in the loop. Technology should invite their involvement and keep them informed.
Every child is different. Voice and conversational experiences need to meet them where they are.
Two billion kids under 13 will grow up talking to AI that responds, adapts, and feels alive. How that goes is being decided right now.
We want to make sure it's built for them.
Let's build something together
Get in touch to learn how UG Labs can power your next project.
Contact Us