Next-day marking and comments on handwritten practice tests for Classes 9–12
TL;DR Nextmark lets a teacher photograph a stack of handwritten practice tests, see suggested marks against her own marking scheme and approve or change them. Every script goes back by the next class with a comment on each answer. Schools and coaching institutes pay per student per year.
Board-year students write tonnes of practice tests. The marked copies come back late, often with just a number and a tick, because one teacher may be checking a few hundred scripts on top of a full timetable. That delay is the real problem.
Feedback is one of the cheapest ways to lift learning. The Education Endowment Foundation's review puts it at about six extra months of progress, and it also warns schools to watch the marking load on staff[1]. NEP 2020 asks Indian schools to move from one-off exams toward regular testing that feeds back into teaching[2]. So who checks 40 scripts across five sections every single week? Nobody, by hand, for long.
What it does, in regulatory terms: marking support for teachers. Nextmark does not certify results or replace board evaluation. Students are mostly under 18, so their scripts are children's personal data under the DPDP Act and Rules, with the school as data fiduciary[3].
The teacher uploads her question paper and marking scheme once. After the test she photographs the scripts on a phone, and Nextmark reads the handwriting in English or Hindi, splits each script by question, proposes step marks and drafts a one-line comment for each answer like "formula right, units missing". She reviews by question and not by script, so she sees all 40 answers to Q7 together and fixes a pattern in one go. Students get their marked scripts the next morning, and the teacher sees which questions most of the class lost marks on.
I reframed the seed, which pitched real-time AI assessment. Speed isn't the main lever - the same review found feedback works whether it comes during, right after or some time after the task[1]. Specific comments and a teacher who stays in charge matter more.
Students and teachers use it but the school or coaching institute pays. I found no published prices from Indian vendors, who sell through demos[4][5], so the starting price is my assumption - a yearly fee per student for Classes 9 to 12, with a free first term. Pitch the principal on teacher time saved and a weekly class-level report. Don't pitch better results, because you cannot promise them.
CBSE marked Class 12 board papers on screen for the first time in 2026, about 98.66 lakh answer books scanned and checked by roughly 70,000 evaluators[6]. Over four lakh students then asked to see their scanned copies[6]. Schools now want students practising for step-marking, and parents are used to seeing a scanned answer book. The DPDP Rules, notified in November 2025, spell out what a vendor handling student data must do[3].
There is no public figure for what Indian schools spend on marking or feedback tools. The narrow anchor is volume. About 24.7 lakh Class 10 and 17.7 lakh Class 12 CBSE candidates sat boards in 2026, across roughly 29,000 affiliated schools[6], and state boards add more on top.
Adjacent - one global grading platform reports 700M+ questions graded at 2,600+ universities[7], which tells you teachers will use photo-based marking once the trust is there.
A Chennai AI-grading startup marks handwritten or digitised answers and cites a rollout to 3,500 students across 18 institutions in one education group[4]. On-screen marking vendors sell OCR-based tools to universities such as Mumbai University, where AI marks a share of sheets after learning from faculty-scored samples[5]. Global photo-grading platforms serve universities abroad[7].
Most of this is aimed at exam evaluation. Weekly practice tests in schools, with comments a student can actually act on, are far less served.
| Difficulty | Moderate (38/100) |
| Starting capital | ₹45–85 L |
| Time to launch | 4–6 months |
Reading messy handwriting, diagrams and Hindi-medium answers from phone photos is the technical work. Trust is the harder part, because one badly marked script shown to a parent can end a school deal and so the review step has to be quick enough that teachers actually do it. School sales also follow the academic calendar, which is slow.
No sector licence applies to marking practice tests. The approvals that matter are about minors' data.
Building this outside India? The categories above usually have a local equivalent, so check with your own regulator.
Don't let AI give final marks without a human check. Messy handwriting, diagrams and Indian-language answers are where it will get things wrong.
Want to work out the AI layer for Nextmark? Talk to Yash →
I like this for a founder who has taught or run academics in a CBSE or state-board school and can win teachers over. If you want to sell outcomes to parents then skip it. Prove two things first - teachers in three schools keep using it weekly for a full term, and they change fewer than one suggested mark in ten.
Plan for roughly ₹45–85 L, covering markers, scanning, software and the first school contracts.
No sector licence. You need a processing contract with each school and DPDP compliance, because the scripts belong to minors.
Yes, wherever schools still test on paper. Check the local data protection law on children's records and match the marking to that country's syllabus.
The free Nextmark validation kit has the line-by-line capital breakdown, the effort scoring, the data-protection checklist for student scripts, a 14-day validation sprint with pass/fail targets, a 90-day launch plan, 10 interview questions for teachers and principals, and 5 assumptions to test. Download the kit →
Thinking about building Nextmark or something close to it? The make-or-break bit is getting marking accuracy high enough that teachers stop double-checking every script, and that's the kind of problem I like working through with founders. 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.