Most AI answering tools know a great deal about engineering in general and nothing at all about your course. Ask one about the specific way your syllabus defines a term, or what Module 4 of your subject actually covers, and it will answer confidently from whatever it trained on — which may not be what your examiner is marking against.
AI Professor works the other way round. It answers from the material you have indexed, and cites the module and page it took the answer from.
Your syllabus documents are read and indexed on the phone. The Sync control shows what it is doing while it works — scanning, reading, extracting text, creating semantic chunks, generating embeddings — with the name of the resource it is on and how many are left.
After that, you ask a question in plain language. The app searches your indexed material, pulls the passages that actually address it, and answers from those.
Every answer drawn from your material carries its source: which resource, which module, which page. You can check it against the document rather than taking it on trust, which is the whole point of citing it.
When the answer is not in your material, the app says so — plainly, before it says anything else — and only then offers a general answer, clearly marked as coming from general knowledge rather than your resources. That distinction is worth more than it sounds. An answer with no material behind it is exactly the answer you should not be revising from, and you should be able to tell the two apart at a glance.
VTU syllabi are not prose. They are headings, module boundaries and prescribed textbook lists, and a document split into arbitrary blocks of a thousand characters cuts straight through the middle of the structure — a chunk that ends halfway through Module 3 and begins inside the textbook list retrieves badly, because neither half is about anything coherent.
The indexing is structure-aware: it recognises the conventions these documents use, splits on module and unit headings, and keeps each section whole. Text without that structure falls back to sentence windows. Each piece keeps its subject, module and page alongside it, which is what makes the citation possible later.
Indexing, searching and answering all run on the device. Your notes are not uploaded to a server to be searched, and there is no account holding a copy of your material. The one thing that leaves the phone is the question you type, which has to reach a model in order to be answered at all.
AI features in the app run on credits. You get a regular allowance, you can earn more by keeping a study streak or inviting a friend, and watching an optional ad tops you up. There is no subscription.
Drop in a PDF, a scanned page or your class notes and ask questions in plain language. Answers come from that material, not from the open internet, so you can trust them against your own syllabus. Text is read on your own phone before anything is sent anywhere.
From the blog
The search was working perfectly. It just never found anything — because truncated embeddings are not unit vectors, and the threshold assumed they were.
Free to download, free to use. Register with an email address or sign in with Google, and your semester is set up in a minute.
