AI-made feelings stories that preschool teachers read aloud, no child screen
TL;DR Feelingtales lets a preschool teacher type in this week's situation (a new sibling, biting, tears at drop-off) and get a reviewed, illustrated five-minute story in the class's language, with picture cards and a note for parents. Preschools pay a yearly licence. The AI only ever writes for adults, and kids just hear a person read.
UNICEF India says over 43% of children under five are at risk of not reaching their developmental potential[1]. Emotional skills are a big part of that, and they get built in conversation. Harvard's Center on the Developing Child describes how naming what a child feels, in back-and-forth exchanges with a caring adult, strengthens the brain circuits behind emotional wellbeing[2].
Preschool teachers know this but have very little to work with.
Most picture books on feelings are in English and set somewhere else. A teacher in Nagpur dealing with three children whose parents just separated has no story that fits, in Marathi, by Monday. How often does this happen? I couldn't find any Indian data on it, so treat it as a hypothesis to test with teachers.
What it does, in regulatory terms: teaching material for adults to read aloud. Feelingtales does not screen, assess, diagnose or treat any emotional or behavioural condition. It never analyses a child's face, voice or mood, and it stores no profile of any child.
Here's how it works - a teacher picks an age band, a language and a situation from a list, or describes one in a line. Feelingtales drafts a short story from templates that an early-childhood specialist and a child psychologist have approved, then adds four picture cards and two questions to ask the class. The teacher reads it aloud and prints the cards. A parent note in the family's language suggests one conversation for that evening.
The seed pitched AI storytelling that talks to the child, and that's been reframed here. A child-facing AI that reads emotions runs into India's data law, which bars tracking or behavioural monitoring of children[3], and WHO advises no more than an hour of screen time a day for children aged 2 to 4[4]. Keeping the AI on the adult's side sidesteps both problems, which is the whole point.
Preschools pay, whether that's a chain buying for every centre or an independent buying per classroom. Teachers use it, and parents get the take-home notes free. The competing price is zero, because one global children's-media nonprofit gives away good emotional-learning material[5]. The test is a yearly licence per classroom (my assumption), sold on time saved and on the parent notes, which preschools can show off to families.
The DPDP Rules were notified on 14 November 2025, with penalties up to ₹200 crore for breaking the children's-data rules[6]. That makes child-facing AI apps riskier and an adult-side tool easier to buy. Language models can now draft a usable story in Hindi, Tamil or Bengali in seconds, which was not true a few years ago.
There is no public figure for spending on emotional-learning material by Indian preschools. There are two adjacent signals - one preschool app includes a social and emotional strand and says more than 5,000,000 parents use it[7], and one activity-kit brand says its kits reach over 10 lakh children[8]. Both show parents pay for early learning, but neither is your market.
One global children's-media nonprofit runs two free programmes that help children aged 4 to 6 recognise and name feelings through stories[5][9]. They're free and well made, but they're fixed. Preschool apps put the child on a screen[7]. Activity-kit brands sell kits and preschool programmes[8]. None lets a teacher make a story for a specific situation in her class's language this week.
| Difficulty | Moderate (37/100) |
| Starting capital | ₹45–80 L |
| Time to launch | 5–8 months |
Generating text is easy. The hard part is quality, because every template needs expert review, every language needs a native editor, and one story with the wrong tone about grief or divorce will cost you a customer. Selling to preschool chains is slow and independents are price-sensitive, so sales won't be a breeze either.
No sector licence for teaching material. The duties come from children's data law and from keeping the product out of clinical territory.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
AI is the core here, but it sits with the teacher, never the child.
Don't trust AI to judge a child's feelings or behaviour. It never sees a child, and anything a teacher reports about distress goes to the school's counsellor.
Want to work out the AI layer for Feelingtales? Talk to Yash →
I like this one for a founder from early-childhood education with a content-editing habit. If you want to build an AI companion for kids then look elsewhere.
Prove one thing first - that 30 teachers in one city use it weekly for two months without being reminded.
Plan for roughly ₹45–80 L, mostly the specialist-reviewed story templates, editing in several languages and selling to preschools.
No sector licence, but India's data law bars tracking or monitoring children and needs parental consent for any child data. Keep the product as teaching material, not therapy or screening.
Yes, if you rebuild the stories for local languages and culture. Check that country's rules on AI and children's data before you sell.
The free Feelingtales validation kit has the line-by-line capital breakdown, the effort scoring, the child-data and content-safety checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for teachers and owners, and 5 assumptions to test. Download the kit →
Thinking about building Feelingtales or something close to it? The make-or-break work is getting teachers to use it every week when free material already exists, and that's the kind of thing I help founders work through. If you need a hand, you know where to find me. Talk to Yash →
For illustration only. Names are fictional and ideas hypothetical; sourced facts are real but may be out of date. Nothing here is legal, financial or professional advice, and Yash Khatri is not personally liable for any use of it. Read the full disclaimer.