Generate accurate multiple-choice questions from any PDF with Synapse. Enhance comprehension and accelerate your learning 🪄

✨ Features

• Uses openrouter, which offers a really wide range of models, and a plethora of API features.
• Beautiful, fast frontend with a slick user experience, powered by Vite and TailwindCSS
• Built in Quiz result to PDF generation powered by React PDF.
• Robust backend powered by ExpressJS, offering a lot of configuration
• Database powered by KeyDB to offer quick retrieval of data.

⚙️ Run It

1. Clone the repository: git clone https://github.com/Syndrizzle/synapse.git
2. Install the dependencies, I recommend using pnpm: 1. Frontend: cd client; pnpm install; 2. Backend: cd ../server; pnpm install
3. Build the frontend: cd ../client; pnpm build
4. ⚠️ IMPORTANT The express server serves the static files from / when you start the server (Later in this process). Copy the dist folder which gets generated after the frontend builds. And move it to server/. Rename it to static. Without this the frontend won't run along with the server.
5. Rename .env.example to .env and fill in your values. You usually just change the OPENROUTER_API_KEY and REDIS_URL as they are a must for it to work.

Important

The Openrouter Key, and the Redis URL is important. Without them, all of this is useless. While you may think openrouter is a paid service, they provide you with a <$1 allowance, thus if you use a cheap model such as Gemini 2.0 Flash, you can easily generate more than 500 quizzes with a free account. Although I have never tested it, I can't say if the <$1 allowance lasts. For redis, you can try Redis Cloud

6. Run the server: node server/index.js
7. If everything goes as planned, the server won't hiccup, and you can visit localhost:3000 to check it out. If you do not see it, check if there's something wrong in the logs.

🚀 Deploy

• You can deploy this on coolify for now. I will work on making it serverless so it can be easier to deploy it on vercel.
• There is a docker-compose.coolify.yml file, use that.
• Open your coolify instance, when creating a new project select Empty Docker Compose and paste it there.
• Optionally configure your desired environment values from .env.example, and the Domain.
• Press deploy to deploy it using docker, the image is hosted on GitHub Container Repository
• For those who do not want to use coolify, a standard barebones docker-compose.yml is also provided.

Happy Exploring 🤠