Poshan AI
Early-warning screening for child malnutrition, built with district administrations and run by Anganwadi workers.
PRODUCTS & IMPACT
What we build, and who it's for.
AI for Indian students. A study companion you can message at 3am, exam notes thousands of students already pay for, and the same technology now catching child malnutrition before it shows.
PRODUCT 01 · COMPANION
Infi Companion. The friend you message at 3am, not a tutor.
A study companion for Class 6 to 12 that talks the way students actually do: Hinglish, casual, on the phone they already use. Ask by chat, by voice, or by snapping a photo of the question. Built gender neutral, with guardrails that redirect with humour instead of shame.
In development · Launching 2026
01
Solve any doubt, any way
Type it, say it, or snap a photo. Infi works through it with you in plain Hinglish.
02
Homework that gets checked
Catch the mistakes before your teacher does, without anyone judging you for them.
03
Every answer ends with a next step
One small nudge per reply, so a single doubt quietly turns into a daily habit.
04
Memory that makes it personal
Infi remembers where you keep getting stuck. Less chatbot, more someone who actually knows you.
PRODUCT 02 · INFINOTES
Exam notes good enough to pay for, cheap enough that anyone can.
Notes for NEET, JEE and Boards in small packs that cost a fraction of a 5,000 rupee coaching subscription. Log in with WhatsApp, pay only for what you need, and ask Infi AI right inside the notes.
Proof, not promises.
The evidence behind everything we build: real students, real revenue.
10K+
Students
12.9%
Free → paid
5 wks
To profitable
35M+
Monthly reach
PRODUCT 03 · POSHAN AI
The same technology, aimed at something bigger than exams.
Poshan AI spots child malnutrition before it is clinically visible and hands Anganwadi workers a simple ranked list of the children to reach first. Its sibling, Angan AI, takes the same engine into Anganwadi systems across Odisha.
AI for public good
Catching what the rules miss.
Against a rule-based baseline, the model caught thousands more at-risk children.
1.47L
Children screened
0.88
Model AUC
9,947
Extra caught vs baseline
Build with us
Want to see where this is headed?
Whether you want to partner, invest, or just get this in front of more students, let's talk.