Halving cost-per-learner on a live-cohort LMS at 4,300 concurrent

Vidyapath Learning runs live, cohort-based upskilling programmes — data engineering, product management, cloud — for working professionals. 40,000 learners on the roll, up to 1,200 concurrent in a single evening class, and a cost problem that was quietly getting worse every month.

They were paying interactive prices for a passive audience

Every live class ran on a general video meeting product, with all 1,200 attendees joined as full participants. But in a lecture, twelve people speak and 1,188 watch. They were paying two-way WebRTC prices, the most expensive way to move video, to deliver a broadcast. The cost per learner for servers rose faster than the price they could charge.

Everything downstream of the class was manual

• Attendance was reconciled by hand from a meeting export, then keyed into a spreadsheet, then into the certification tracker. For a cohort model where attendance drives certification eligibility, this was the most important data in the business and also the least trusted.
• The question bank had leaked. Their test tool sent the full paper to the browser when an exam started. A learner with the browser devtools open could read every question — and the answers, in one memorable case — before answering the first one. Screenshots of the bank were circulating in cohort WhatsApp groups.
• Proctoring cost more than it was worth. Their vendor recorded the full session and charged for a person to review all of it. They were paying to have a person watch 40 minutes of video of someone doing nothing suspicious, thousands of times a month.

And it all happened at once

Traffic is not spread across the day. It comes as a spike. Working professionals attend class between 8pm and 10pm IST. The platform was idle for twenty hours a day and under heavy load for four, and it was built for that peak.

Moved one group at a time over a term. The old and new classes ran side by side for three weeks, so learners could go back if the new one was worse. After the second week, nobody switched back.

Results

• Infrastructure cost per learner fell 58%. The overwhelming majority is the WebRTC-to-HLS split. This was not a small tweak; it corrected a basic mistake about what a lecture is.
• Peak concurrency reached 4,300 without incident — more than triple the load that used to break them — during a free masterclass they ran because they finally trusted the platform to take it.
• Proctoring cost per exam dropped 71%. Same integrity outcomes, a fraction of the human review hours, because people now only check where a check is needed.
• Zero question-bank leaks since launch, across nine assessment cycles. The bank cannot leak from a client that is never sent it.

What it did to the business

Attendance data went from a spreadsheet somebody maintained to a stream the product reads, and the certificate process stopped being a monthly manual task. More importantly, the cost picture flipped: adding learners to a cohort now costs them very little, so each extra cohort now makes money, which it clearly did not before.

They have since doubled cohort sizes on their two largest programmes without adding infrastructure spend, which is the outcome that turns an engineering project into a business one.