Week of 25 May
Watching whether children re-read a word more often when they got it right the first time.
Every child is born curious.
The question is what they learn to do with that curiosity, and who they trust to help them. For the first time in history, part of the answer to that question is going to be a machine. How that machine behaves — what it pays attention to, what it ignores, what it is patient with — will shape a generation of children.
Childhood is not an optimization problem.
We are a small team building software for children and the parents raising them. We watch carefully, build slowly, and try to earn trust over time. We take play seriously, because play is how children do their most serious thinking. We take parents seriously, because parents are partners, not users.
Many children in India grow up speaking one language at home, another at school, and increasingly a third through screens. Most AI was not built for them.
This is the place we are building from, and the children we are building for. Multilingual, mobile-first, surrounded by family, learning a language that is not their first — and meeting AI before they meet a teacher. We think this child deserves software designed for them specifically, not adapted from elsewhere.
Week of 25 May
Watching whether children re-read a word more often when they got it right the first time.
Week of 18 May
A question we're sitting with: does encouragement work better before a child stumbles, or after?
Week of 11 May
Studying how multilingual children switch languages while sounding out unfamiliar words. The transitions are faster than we expected — almost always away from the family language.
Reads with children — patiently. ZigZu listens as a child reads aloud, helping gently along the way — the way a good teacher would.
See ZigZuHelps families make sense of raising a child. ParentMind organizes the advice, notes, reels, and questions that modern parenting leaves scattered everywhere.
See ParentMindWe are a small team. We do not run focus groups or design sprints. We watch children read, test, and play — usually in their own homes, often with their parents and grandparents in the room. We take notes. We build prototypes. We try them again the next week.
We test by playing. We measure by watching. We ship when a child reaches for it twice.
We treat the things we believe as commitments, not slogans. They are written down and they are public — so they can be held against us when we get something wrong.
I started ANA PlayLabs to spend the rest of my career on a single question: how can AI grow up well alongside children — and help, rather than crowd out, the people raising them?
Anshul
Founder · Previously at Airtel, OYO, and Network18 · Building ANA with AI leverage
We believe childhood deserves patience, privacy, and human judgment.
We do not sell children's data. We do not design for addiction. We do not believe every moment of childhood should happen on a screen.
We build slowly, test carefully, and try to earn trust over time.
The next essays are being written. More to come.
If you want to follow what we're building, leave us your email.
Or write to us: hello@anaplaylabs.com