Edit description, writeup (Markdown), and notes for each concept across all topics.
Walk the bank one row at a time. Enable keeps the question live. Disable flips is_suspect=1 so it stops appearing in tests, practice, and /q/ views. Use the filter to resume an earlier pass.
Each AI tutor has a unique personality and system prompt that controls how it responds to students. Edit the profile to change the AI's behavior.
?utm_source=telegram&utm_medium=bot&utm_campaign=jee2027
Uses advanced multi-modal AI. Generated post is saved as draft — review and publish from the blog list.
Import videos from your YouTube playlist into the concept videos library.
Upload a video file directly to your YouTube channel and auto-add it to the library.
Configure the AI backend for chat tutors. Choose a provider, enter your API key, and select a model. Without a key, tutors use pre-written fallback responses.
Vision is used when a student attaches a photo to a chat ("explain this circuit") or uploads a short clip ("why is this experiment going wrong?"). Most modern chat models also handle vision; pick a different model here only if you want to use a cheaper or stronger one for analysis specifically.
Automatically back up the database to Amazon S3 (or any S3-compatible storage like MinIO, DigitalOcean Spaces, Backblaze B2). Backups include all user data, test attempts, chat history, and settings.
Errors and issues reported by users from concept, question, video, and topic pages.