Restore Dockerfile and README for HF Spaces deployment

Dockerfile

@@ -26,15 +26,15 @@ ENV LLM_PROVIDER=huggingface
 
 # Create startup script
 RUN echo '#!/bin/bash\n\
-# Start FastAPI backend in background\n\
-uvicorn backend.main:app --host 0.0.0.0 --port 8000 &\n\
-\n\
-# Wait for backend to start\n\
-sleep 2\n\
-\n\
-# Start Streamlit frontend\n\
-streamlit run app.py --server.port 8501 --server.address 0.0.0.0 --server.headless true\n\
-' > /app/start.sh && chmod +x /app/start.sh
+    # Start FastAPI backend in background\n\
+    uvicorn backend.main:app --host 0.0.0.0 --port 8000 &\n\
+    \n\
+    # Wait for backend to start\n\
+    sleep 2\n\
+    \n\
+    # Start Streamlit frontend\n\
+    streamlit run app.py --server.port 8501 --server.address 0.0.0.0 --server.headless true\n\
+    ' > /app/start.sh && chmod +x /app/start.sh
 
 # Run startup script
 CMD ["/app/start.sh"]

README.md

@@ -1,5 +1,6 @@
 ---
 title: FocusFlow
+emoji: 📚
 colorFrom: blue
 colorTo: purple
 sdk: docker
@@ -7,52 +8,10 @@ app_port: 8501
 pinned: false
 ---
 
-# FocusFlow - AI Study Companion
+# 📚 FocusFlow - AI Study Companion
 
-An intelligent study assistant powered by AI that transforms your learning materials into personalized, adaptive study experiences.
+Intelligent study assistant with RAG-powered Q&A, adaptive quizzes, and personalized learning paths.
 
-## Features
+**Live Demo**: Running on Hugging Face Spaces with Llama-3-8B
 
-- ** Multi-Subject Study Planning**: Upload PDFs and get automated multi-day study plans
-- ** RAG-Powered Q&A**: Ask questions and get answers with source citations
-- ** Adaptive Quizzes**: Context-based quizzes that adapt to your performance
-- ** Progress Tracking**: Track mastery levels and quiz history
-- ** Smart Day Progression**: Automatically unlocks new topics as you complete them
-- ** Source Citations**: Every answer cites the exact source and page number
-
-##  Models Used
-
-- **LLM**: Meta-Llama-3-8B-Instruct (via Hugging Face Inference API)
-- **Embeddings**: nomic-embed-text (for semantic search)
-
-##  How to Use
-
-1. **Upload PDFs**: Add your study materials in the Sources panel
-2. **Generate Plan**: Ask the Calendar to create a study plan (e.g., "Make a 5-day plan")
-3. **Study**: Click on topics to view lessons and ask questions
-4. **Take Quizzes**: Test your knowledge and unlock new topics
-
-##  Demo Note
-
-This is a **cloud demo version** running on Hugging Face Spaces using the Llama-3-8B model.
-
-**For offline/local use** with enhanced privacy and llama3.2:1b (no internet required):
--  [GitHub Repository](https://github.com/thesivarohith/hack)
--  [Local Setup Guide](https://github.com/thesivarohith/hack/blob/main/RUN_GUIDE.md)
-
-The local version works completely offline and keeps all your data private on your machine.
-
-##  Tech Stack
-
-- **Frontend**: Streamlit
-- **Backend**: FastAPI + LangChain
-- **Vector DB**: ChromaDB
-- **LLM**: Hugging Face Inference API
-
-##  License
-
-MIT License - See [LICENSE](https://github.com/thesivarohith/hack/blob/main/LICENSE) for details.
-
----
-
-Built with ❤️ for better learning experiences
+**Local Setup**: [GitHub Repository](https://github.com/thesivarohith/hack)