LectureLens

An AI-powered study assistant that helps students understand their lecture materials. Upload lecture PDFs or text files, get instant summaries, extract key concepts, and chat with an AI tutor grounded in your lecture content.

Live: Deployed on Cloudflare Workers + Pages Link: https://lectrurelens.pages.dev

---

Features

• Upload Lectures — Upload PDF or TXT files (up to 50 MB). Content is extracted, stored, and indexed per user.
• AI Summarization — Generate structured, Markdown-formatted summaries. Long documents are automatically chunked, summarized in parts, and merged into a cohesive overview.
• Concept Extraction — Extract key definitions, formulas, and core theoretical concepts from any lecture.
• Contextual Chat — Ask questions in plain English and get answers grounded in your uploaded lecture content, powered by a persistent chat history.
• Authentication — Email/password signup with PBKDF2 hashing, plus Google Sign-In (OAuth). Session-based auth with Bearer tokens.
• Rate Limiting — Per-user and per-IP rate limiting via Durable Objects to prevent abuse.
• Lecture Management — View, select, and delete your uploaded lectures from a sidebar.

---

Tech Stack

Layer	Technology
Frontend	Vanilla HTML/CSS/JS, served via Cloudflare Pages
Backend	Cloudflare Workers (TypeScript)
AI Model	Cloudflare Workers AI — @cf/meta/llama-3.3-70b-instruct-fp8-fast
Database	Cloudflare D1 (SQLite) — users, sessions, user-lecture mappings
Stateful Storage	Cloudflare Durable Objects — lecture content & chat history (LectureMemory), rate limits (RateLimiter)
Auth	PBKDF2 password hashing + Google Identity Services
PDF Parsing	PDF.js (client-side extraction)

---

Project Structure

LectureLens/
├── frontend/                   # Cloudflare Pages frontend
│   ├── index.html              # Landing page, auth forms, main app UI
│   ├── style.css               # Styling
│   ├── script.js               # Client-side logic (auth, upload, chat, sidebar)
│   └── functions/api/          # Pages Functions (proxy to Worker backend)
│       ├── [[path]].js         # Catch-all proxy for /api/*
│       └── auth/
│           ├── login.js
│           └── signup.js
├── worker-backend/             # Cloudflare Worker backend
│   ├── src/
│   │   ├── index.ts            # Main Worker entry — all API route handlers
│   │   ├── auth.ts             # Password hashing (PBKDF2) & session validation
│   │   ├── LectureMemory.ts    # Durable Object for lecture storage & chat
│   │   └── RateLimiter.ts      # Durable Object for per-endpoint rate limiting
│   ├── migrations/
│   │   ├── 0001_add_lecture_metadata.sql
│   │   └── 0002_add_google_auth.sql
│   ├── wrangler.jsonc          # Wrangler config (bindings, D1, DOs, AI)
│   └── package.json
├── schema.sql                  # Base database schema (users, user_lectures, sessions)
└── README.md

---

API Endpoints

Authentication

Method	Path	Description
POST	/api/auth/signup	Register with email + password
POST	/api/auth/login	Login with email + password
POST	/api/auth/google	Sign in / sign up with Google
POST	/api/auth/logout	Invalidate the current session

Lectures

Method	Path	Description
POST	/api/upload	Upload a lecture file (multipart/form-data)
GET	/api/my-lectures	List all lectures for the authenticated user
DELETE	/api/lectures/:id	Delete a lecture

AI Features

Method	Path	Description
POST	/api/chat/:lectureId	Send a chat message about a specific lecture
POST	/api/summarize	Summarize lecture text
POST	/api/extract-concepts	Extract key concepts from a lecture

Other

Method	Path	Description
GET	/api/stats	Public aggregate stats (user count, lecture count)
GET	/	Health check

All authenticated endpoints require an Authorization: Bearer <token> header.

---

Database Schema

users — User accounts (email/password or Google OAuth)

sessions — Auth session tokens with expiration

user_lectures — Maps users to their uploaded lectures (with name and timestamp)

---

Rate Limits

Endpoint	Limit	Window
Chat	15 requests	per minute
Summarize	5 requests	per hour
Extract Concepts	5 requests	per hour
Upload	10 requests	per hour
Signup	3 requests	per hour (per IP)
Login	10 requests	per hour (per IP)

---

Getting Started

Prerequisites

• Node.js (v18+)
• Wrangler CLI (npm install -g wrangler)
• A Cloudflare account with Workers, D1, Durable Objects, and Workers AI enabled

Setup

1. Clone the repository

   git clone https://github.com/<your-username>/LectureLens.git
   cd LectureLens

2. Install backend dependencies

   cd worker-backend
   npm install

3. Create the D1 database

   wrangler d1 create lecturelens_db

   Update the database_id in worker-backend/wrangler.jsonc with the ID returned.

4. Run database migrations

   wrangler d1 execute lecturelens_db --file=../schema.sql
   wrangler d1 execute lecturelens_db --file=migrations/0001_add_lecture_metadata.sql
   wrangler d1 execute lecturelens_db --file=migrations/0002_add_google_auth.sql

5. Set secrets (for Google OAuth)

   The Google Client ID is already configured in wrangler.jsonc. If using your own Google OAuth credentials, update the GOOGLE_CLIENT_ID var.

6. Run the backend locally

   npm run dev

7. Serve the frontend

   In a separate terminal, from the project root:

   cd frontend
   npx wrangler pages dev . --port 8788

---

Deployment

Backend (Worker)

cd worker-backend
npm run deploy

Frontend (Pages)

Deploy the frontend/ directory to Cloudflare Pages via the dashboard or CLI. The Pages Functions in frontend/functions/ proxy /api/* requests to the Worker backend.

---

License

This project is for educational purposes.