Parent-filmed play clips that help paediatricians spot autism signs sooner
TL;DR Sproutreel gets parents of toddlers aged 18 months to 5 years to film three short, guided play clips at home. Trained developmental reviewers score them against a checklist and the child's paediatrician gets a screening summary. Clinics pay per case, and it works if worried families reach assessment months sooner.
In a study across five Indian districts, 12.0% of children aged 2–9 had at least one neurodevelopmental disorder, and 9.2% of those aged 2 to under 6[1]. Catching these early matters a lot, but a paediatrician sees a toddler for a few minutes and the kid is usually crying or asleep.
Government help does exist. District Early Intervention Centres under the RBSK programme take referrals, but a 2024 study of one centre recorded staff shortages and heavy caseloads, and it cites earlier work in which 73.4% of parents said lost wages kept them from repeat visits[2].
So families wait. And the window when early support helps most keeps shrinking.
What it does, in regulatory terms: developmental screening support for a doctor. It does not diagnose autism or any other condition. A trained human reviewer scores every case and the paediatrician decides on referral. Software that diagnoses would likely be a medical device under the Medical Devices Rules 2017[7], so Sproutreel stays firmly on the screening side.
The app tells parents what to film - calling the child's name, a shared toy game and a pointing task, each under two minutes. Software checks that each clip is usable and marks the moments a reviewer needs to see. A reviewer, trained and supervised by a developmental paediatrician, scores the clips against a standard checklist and writes a short summary for the doctor.
The seed promised AI video analysis of child development, and I've moved the AI back a step. An automated autism label would be a diagnostic claim with heavy evidence needs, and a false alarm frightens parents half to death. Software helps with clip quality and speed while people do the judging.
Parents film but the paediatric clinic or children's hospital buys. Clinics can charge families for a screening add-on. I found no Indian price for video screening, so my rough assumption is ₹1,500–3,000 per case. Therapy chains are a second buyer, since a clear screening summary fills their assessment slots with the right families.
Investors are backing child development services. One autism-therapy chain raised ₹65 crore in August 2026 and runs about 90 centres[5]. The DPDP Act now makes verifiable parental consent compulsory for children's data[6], which suits a service built around consent from day one. In the US, one autism diagnosis aid is authorised by the FDA for children aged 1.5 to 6[3], so regulators will accept remote evaluation.
There's no public figure for what India spends on developmental screening. The narrow figure is need - 9.2% of children aged 2 to under 6 had a neurodevelopmental disorder in the five-district study[1].
The adjacent signal is that therapy chains are raising money and opening centres[5], and they'll need better-screened referrals to fill them.
One Indian autism-care company offers a free in-app questionnaire for children aged 1.5 to 5 and runs therapy centres in more than 12 locations[4]. Centre-based therapy chains provide therapy and assessment. A US diagnosis aid goes further and supports a diagnosis[3].
What's missing is a video-based screen that fits a paediatrician's normal day, before a family ever reaches a therapy centre.
| Difficulty | Moderate (47/100) |
| Starting capital | ₹54–95 L |
| Time to launch | 6–9 months |
Filming instructions that work in small, noisy homes will take many rounds of testing. Training reviewers to score consistently and then checking them against expert assessments is slow work, and there's no shortcut. Consent and secure storage for children's videos have to be right from the first day.
No device licence if it stays a screening service with human reviewers. The children's data rules are the strict part.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Never let AI label a child as autistic or as fine. A false alarm scares parents and false reassurance delays help, so trained people score every case and the paediatrician decides on referral.
Want to work out the AI layer for Sproutreel? Talk to Yash →
Here's what can go wrong -
I like this for a founder working with a developmental paediatrician who'll own the clinical standard. If you're planning an AI that labels children then it is not for you. First prove, with 100 children, that your screens agree with full expert assessments and that clinics send more families to assessment.
Plan for roughly ₹54–95 L, mostly reviewer training, the app and secure storage for children's videos.
Not a device licence if humans do the scoring. You need verifiable parental consent under the DPDP Act, and doctors advising from the review follow the telemedicine guidelines.
Yes, though rules vary. In the US, one autism diagnosis aid is FDA-authorised because it supports diagnosis, so check your own regulator early.
The free Sproutreel validation kit has the line-by-line capital breakdown, the effort scoring, the consent and data-protection checklist, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for paediatricians and parents, and 5 assumptions to test. Download the kit →
Thinking about building Sproutreel or something close to it? The make-or-break work is proving your reviewers' screens match expert assessments, and that's the kind of thing I help founders figure out. 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.