Everyscheduledclass,verified.
Colleges already run an academic audit — attendance sheets, contact-hour reports, syllabus compliance. Today it is assembled by hand, once a semester, from paper. We produce the same record every day, automatically, from the cameras the campus already owns.
Were all of yesterday's scheduled classes actually held?
Ask any principal in the country and the honest answer is that nobody knows. The register says one thing, the timetable says another, and the truth sits in rooms nobody had time to walk past. By the time it matters — an accreditation visit, an exam-eligibility dispute, a parent asking a fair question — the evidence has to be reassembled from memory.
This is not a surveillance product. It is an accountability record that happens to run on cameras. The subject is the scheduled class, not the student. Every figure traces back to the moment it came from, so a disputed number can be answered rather than argued about.
Minimum attendance every institution has to evidence
UGC / AICTE, with 10% condonation
The NAAC section that asks for teaching-delivery evidence
Teaching, Learning & Evaluation
Cameras campuses are required to have and rarely use
UGC anti-ragging directive
Everyone sees the same truth, scoped to their job.
The teacher, the head of department and the principal all read the same underlying record. What differs is how much of it they see and what they are asked to do about it.
For the teacher
Your own record, before anyone else acts on it. What you taught, how many were in the room, and how your contact hours are tracking against the syllabus. If something looks wrong, you add the reason and it closes there.
For the head of department
The handful of classes that did not go to plan today, and the recurring problems worth raising this week. Delivery facts only — no teaching-quality score, no ranking of your faculty.
For the principal
One number for whether teaching is happening as planned, the exceptions that need an owner, and accreditation evidence in the format the assessor asks for — ready, not reconstructed at the last minute.
Open it and click around.
This is the working prototype, not a screenshot. Switch between the eight roles in the top right — teacher, head of department, the deans, the registrar, the director, and a read-only auditor who cannot change anything at all — change the size of the institution, and click any cell in the grid to see the evidence behind that verdict. It reads in English, Marathi or Hindi, in light or dark. Every figure on every screen is illustrative — no real campus data has been processed.
Built so students and staff defend it, not resist it.
A system a campus cannot explain in plain language to its own students is a system that gets switched off within a term. These limits are design decisions, written down before anything was built.
No camera identifies anyone
The cameras count people. They do not recognise faces, keep a gallery, or name anyone — and cannot be configured to. Where the system does hold student names, in marks and registers, they come from records the college already keeps on paper.
No emotion or mood reading
Not built, anywhere. It has no established scientific basis and is banned in European classrooms.
No continuous recording
A single still image every five minutes. There is no video to watch.
Nothing leaves the campus
Everything is worked out on a box in your own server room. Images are discarded once counted.
The people who run a campus are not at a desk.
Heads of department are in corridors, technicians are up ladders, the principal is walking the building. A companion app carries the handful of things that cannot wait — and puts a QR sticker on every classroom door, so staff attendance becomes a record rather than a guess.