2026–27 Founding pilot open to two schools
A personal AI tutor for every child. Even offline.
Enumora teaches your school’s syllabus in the child’s own language, on the Android phone the family already owns. It listens to how a child reasons, finds the misconception, and answers with a question — never the answer. Teachers stay in command, and every mastered topic ends in a credential anyone can verify.
- Works with no internet
- Tutor speaks Hindi and English
- Conversations never leave the phone
A real lesson script from the product. Step through it, switch to English, or open the Pedagogy view to see why each turn exists.
Why we exist
Parents pay for tuition because one teacher cannot reach forty children.
Great teaching hasn’t reached everyone, and today’s cloud-first, English-first education technology was never built for the child who needs it most. Four numbers define the problem.
cannot read a Class-2 text (ASER 2024). The gap is not ability. Nothing was built for how, where, and in which language these children learn.
is what families already spend on private tuition to buy the personal attention a crowded classroom can’t give — ₹6,000 to ₹24,000 a year, per child.
Personal attention is arithmetically impossible. A teacher who wants to reach every child needs help that works on her schedule, not around it.
Most of India’s scheduled languages have no production-grade AI tutor. A child should be able to ask why plants need light in the language she thinks in.
Sources: ASER 2024 (Pratham); household tuition spend from Enumora research; language coverage from Enumora analysis of public product documentation, September 2026.
How Enumora is different
Every AI chatbot assumes a child with a device, the internet and the urge to ask. Most children are not that child.
So we didn’t build a smarter chat window. We built school infrastructure: tutoring that arrives on the school’s schedule, teaches the school’s syllabus in the child’s language, keeps the teacher in command, works where the internet doesn’t, and ends in proof.
Offline at the point of use
The tutor runs on the phone itself. Dialogue, quizzes, diagrams and voice are generated on the device — at 8 pm, in a power cut, with no data pack.
Never hands over the answer
A runtime rule, not a promise: any tutor turn that would leak the answer is withheld. The child does the thinking — that is what tutoring is for.
The teacher is in command
She schedules sessions, sees where the class is stuck, and pushes re-teach sets. The pulse reports the class to her — never her to anyone.
Ends in proof, not screen time
Mastery earns a signed credential that anyone can verify in under a second. Nothing in the product rewards minutes.
How it works
One day with Yuvaan, Class 7, Agra.
8:00 in the morning
Yuvaan picks up his mother’s phone at home. There is no signal, and there doesn’t need to be. He asks, in Hindi, why plants need light — could a candle work? The tutor answers with a question of its own, listens to how he reasons, and adjusts.
11:00 in class
His teacher sees where the class is stuck and pushes a re-teach set. She sees class-level patterns — not Yuvaan’s conversations. That isn’t a policy promise. The data that syncs to school has no field to carry them.
5:00 in the evening
Yuvaan secures the unit. A signed credential — Photosynthesis, 84% mastery — lands on his phone, co-signed by his teacher. His parents see it at the next PTM. Years from now, an employer can verify it in a second.
Imagine every child with a private tutor who knows their language, their textbook and their pace — and who never gives up on them. That is what we are building.
What we’re building
One platform. Every layer a learner needs.
Each layer is strong alone and unmatched together — the tutor, the device it runs on, the proof it produces, and the research that will shape what comes next.
enum enumora.platform {
{e}.brain In build
The on-device tutor: a compact language model with speech recognition and a warm voice, running a curriculum-aligned lesson engine. Socratic by design — hint ladders and counter-questions, never bare answers.
{e}.edge For government schools
The device surface. In private schools, Enumora runs on the Android phones families already own. For government tenders that need device and platform from one vendor, {e}.edge supplies a rugged tablet with the tutor built in.
{e}.pass In build
Credentials signed on the device at the moment of mastery, endorsed by the school on sync, and verifiable by anyone in under a second — even years later.
{e}.sense In research
Attention-aware tutoring research: helping the tutor slow down, offer a hint earlier, or suggest a break. Off by default, parent-managed, and shipped only after peer review.
{e}.labs In research
AI-instrumented vocational training for India’s ITIs: cameras and sensors on standard workshop equipment, giving every trainee the specific feedback a master craftsman would.
} // see it in action
Who it’s for
Sold to schools. Built around the people in them.
For schools
An admissions argument parents already understand: a personal tutor for every child, included in the fees — with outcomes you can show at every PTM.
For teachers
One-to-one attention for all forty children, on your schedule. The tutor takes the chores; the mastery belongs to your classroom.
For parents
Patient, Socratic help at home in your child’s language — on your own phone, with no data pack, and with conversations that stay private.
Trust & privacy
There is no setting to share a child’s data. The screen to turn it on does not exist.
A child’s conversations with the tutor never leave the device. Not encrypted-and-sent — never sent. There is no server holding them, so there is nothing to leak, subpoena or sell.
- Consent child by child. One sibling’s consent never covers another. A child without consent still learns; their data goes nowhere.
- Outcomes cross, process never. Schools see mastery and where misconceptions cluster — never what a child asked or where she hesitated.
- Mastery, not minutes. We count learning occasions and secured topics. A tutor should make itself less necessary.
- Designed for India’s DPDP Act from the first line of code, then built stricter.
Questions
The questions schools ask us first.
Honest answers, including the ones about what Enumora can’t do yet.
Students can use free AI chatbots. Why does our school need this?
Chatbots serve the child who owns a device, has the internet and opens the app to ask. Enumora arrives on the school’s schedule, teaches the school’s syllabus in the child’s language, works without the internet, keeps the teacher in command and ends in verifiable proof of mastery. A chat window is a library; this is a school.
Is it really offline?
The teaching is. Conversation, quizzes, diagrams and voice are generated on the phone with no connection. Video is not generated on a phone — no honest product can claim that — so lesson videos arrive in content packs whenever the phone next meets Wi-Fi.
Will it replace teachers?
No, and it is built so it cannot. The teacher schedules the tutor, the class pulse reports to her alone, and every credential carries her co-sign.
What does it cost?
The pilot term is free, with success criteria agreed in writing first. After that, a per-student annual licence of around ₹1,499 (about ₹125 a month), finalised with our founding schools.
Ask us for a demonstration. Then switch off the Wi-Fi.
That is the whole pitch. A pilot conversation takes thirty minutes, and the founding cohort for 2026–27 is limited to two schools.