The situation
The current collection is 33 sets of textbook files with English and Marathi text, 461 chapters in all, with Hindi text in 19 of the sets. On paper, that covers a lot of Classes 5 to 10 of the Maharashtra State Board.
The problem
Book files are not tidy. Some still have text errors. Some do not separate chapters correctly, which means the app can look in the wrong place. And not every subject or book is in the collection. A family whose child uses a book we do not have gets no help at all, however good the AI is.
What we tried
We fixed files one at a time and kept a record of which still need work. We made the app search the book text first, so a reply begins from the chapter rather than from general knowledge. And we stopped the website from listing subject names it cannot verify.
What we learned
The honest number is not 461. It is “461, some of which still need fixes”. The honest answer about which books are included is not “all your books”. It is “check the exact book you need”.
What it changed
The books and languages page shows the collection as a table with its limits in the same row. A space on that page is kept for a real screenshot of the screen where you pick your class and book, so the website can show the actual list rather than describe it.
Takeaway
Before anything else, tell us the textbook names. That one detail does more to predict whether VIDYA AI will help than anything about the AI.
The numbers here come from our release reports and one reported phone test. They do not show any change in marks or exam results.

