Content Management
Upload study materials for your assigned batches — course material, lecture notes, assessments, practice problems, and reference notes. Uploaded PDFs are automatically processed for AI-powered features, and you can extract questions from them to build your question bank.
Video Walkthrough
Overview
Navigate to Academic Management → Batch Content from the sidebar. The page shows your assigned batch and all uploaded content files.
Batch Selector
Your assigned batch is pre-selected at the top. If you teach multiple batches, use the dropdown to switch between them.
Uploading Content
Click the + Upload button to add new study material:
- Select the content category:
- Lecture Notes — classroom teaching material
- Practice Material — problem sets and exercises
- Reference Notes — supplementary reading
- Assessments — test papers and sample exams
- Choose the subject (Physics, Chemistry, Mathematics, etc.)
- Select a PDF file from your computer
- Click Upload — processing begins automatically
Automatic Processing Pipeline
After upload, the system automatically:
- Extracts text from the PDF
- Chunks the content into paragraphs and sections
- Embeds each chunk as a vector using AI models
- Marks the file as Ready
Once ready, the content is searchable by students through the AI Doubt Solver and Smart Tutor.
Content Table
Each uploaded file shows:
| Column | Description |
|---|---|
| Category | Content type — Lecture Notes, Practice Material, etc. |
| File | File name with PDF icon |
| Subjects | Subject tags (Physics, Chemistry, etc.) |
| Size | File size |
| Status | Processing status — Ready (fully processed), Processing, or Failed |
| Uploaded | Upload date and uploader email |
| Actions | Preview, Download, Extract Qs, Re-embed, Delete |
Use the subject filter dropdown to narrow the list.
Extracting Questions from Content
One of the most powerful features available to teachers is the ability to extract questions directly from uploaded study materials.
Click the Extract Qs button on any uploaded file with "Ready" status:
- AI reads the PDF content
- Identifies question patterns (MCQs, problem statements, etc.)
- Extracts questions with options, correct answers, and explanations
- Adds them to the Question Bank as AI-extracted questions
The number of extracted questions appears as a badge on the file (e.g., "20 Qs").
After extraction, navigate to the Question Bank to review the extracted questions. Verify accuracy, adjust difficulty levels, and publish them for use in quizzes.
Other Content Actions
| Action | What it does |
|---|---|
| Preview | Open the PDF in a preview modal to view the content |
| Download | Download the original file |
| Re-embed | Re-process the file's vector embeddings (useful if the embedding model has been updated) |
| Delete | Remove the file and all associated chunks, embeddings, and extracted questions |
Review Queue
Click the Review Queue tab to see content files that need review or have processing issues.
How Content Powers AI Features
Uploaded content is used across multiple AI-powered features:
| Feature | How content is used |
|---|---|
| AI Doubt Solver | When a student asks a doubt, the AI searches your uploaded content for relevant context and cites it in the answer with page references |
| Smart Tutor | The AI Tutor references your batch content during guided learning sessions to provide accurate, curriculum-aligned explanations |
| Question Extraction | AI extracts MCQ questions from your PDFs to rapidly build the question bank |
| RAG Context | All content chunks are available as retrieval-augmented generation (RAG) context for AI responses |
Content Visibility
| Scope | Who can see |
|---|---|
| Batch content | Students enrolled in that specific batch only |
| Program content | All students in the program (uploaded by org admin) |
| Public content | All students platform-wide (uploaded by system admin) |
Upload separate PDFs for each chapter rather than one large file. This improves chunk quality and makes AI citations more precise — students see specific page references instead of generic file-level citations. It also produces better results when extracting questions.