SmartPDF: Summarize PDFs with Llama 3.3

SmartPDF leverages Together AI to power Llama 3.3, delivering rapid PDF summaries organized by chapter.

Technical Stack

• Inference: Together AI
• LLM: Llama 3.3
• Framework & Tools: Built with Next.js, Tailwind CSS, and TypeScript. Prisma ORM interfaces with a Neon (PostgreSQL) database. Observability is managed via Helicone, while Plausible handles analytics. PDF files are stored in Amazon S3.

Local Setup

1. Clone the repository: git clone https://github.com/Nutlope/smartpdfs

2. Configure environment variables. Create a .env file using .example.env as a template. You will need to provide the following:

   • TOGETHER_API_KEY=
   • DATABASE_URL=
   • S3_UPLOAD_KEY=
   • S3_UPLOAD_SECRET=
   • S3_UPLOAD_BUCKET=
   • S3_UPLOAD_REGION=us-east-1
   • HELICONE_API_KEY= (Optional: for observability)

3. Install dependencies: pnpm install

4. Generate the Prisma client: pnpm prisma generate

5. Launch the development server: pnpm dev

Project Roadmap

• Implement IP-based rate limiting
• Integrate OCR support to parse images within PDFs
• Add a revision step to further refine summary quality