Adobe Round 3 Challenge

This project is a multi-service application consisting of:

• A Next.js frontend (nextServices/)
• A FastAPI backend (pythonServices/)
• A MongoDB database

All services are orchestrated using Docker Compose for easy setup and deployment.

Folder Structure

adobe-r3/
│   .dockerignore
│   .gitignore
│   Dockerfile
│   entrypoint.sh
│   nginx.conf
│   supervisord.conf
├── nextServices/
│   ├── .dockerignore
│   ├── components.json
│   ├── next-env.d.ts
│   ├── next.config.mjs
│   ├── package.json
│   ├── pnpm-lock.yaml
│   ├── postcss.config.mjs
│   ├── tailwind.config.ts
│   ├── tsconfig.json
│   ├── app/
│   ├── components/
│   ├── contexts/
│   ├── hooks/
│   ├── lib/
│   ├── public/
│   ├── scripts/
│   ├── styles/
│   └── uploads/
├── pythonServices/
│   ├── .dockerignore
│   ├── requirements.txt
│   ├── Dockerfile
│   ├── server.py
│   ├── ... (other Python scripts and model files)
│   ├── chroma_storage/
│   ├── saved_models/

How to Run

1. Clone the repository

2. Build Docker

docker build -t adobe-round3 .

3. Run container

Include all the environment variables with -e flag. All the ports must be included as given below:

docker run -it -p 8080:8080 -p 8000:8000 -p 27017:27017 --name adobe-r3-container -e ADOBE_EMBED_API_KEY="...." -e GOOGLE_API_KEY="...." -e TTS_PROVIDER="..." -e GEMINI_MODEL="gemini-2.5-flash" -e LLM_PROVIDER="gemini" -e AZURE_TTS_ENDPOINT="...." -e AZURE_TTS_KEY="....."

Features

• PDF Upload & Folder Organization
  Upload PDFs and keep them neatly grouped by folders for better management and quick access.

• User Authentication
  Secure user login to manage and access personalized document collections.

• Smart PDF Store

  • Automatic PDF heading detection and structured chunking using Chroma DB.
  • AI-powered PDF summaries and FAQ generation for faster understanding.

• AI-Powered Podcasts & Reading Guides

  • Generate a 3-minute podcast summarizing all PDFs in a folder.
  • Get an intelligent reading guide that recommends an optimal reading order based on document summaries.

• Interactive Document Navigation

  • Navigate through PDFs via detected headings.
  • Select any text and retrieve related content from the same document or other PDFs in the folder.

• Insight Generation with LLM

  • Select text passages and generate AI-driven insights.
  • Discover supporting information, contradicting viewpoints, and fun “Did You Know?” facts to enrich understanding.