requirements.txt

gunicorn
# Swift Check AI Requirements
huggingface-hub>=0.23

# --- Core API framework ---
flask>=3.0.0                        # Your existing routes

# --- LLM & RAG ---
groq>=0.24.0                        # Groq client for LLM
langchain>=0.2.1                    # RetrievalQA, document loaders, etc.
langchain-community>=0.2.0          # Community integrations for ChromaDB

# --- Vector database ---
chromadb>=0.5.0                     # Lightweight, persistent vector store
sentence-transformers>=2.7.0        # Local embeddings (all-MiniLM-L6-v2)
huggingface-hub>=0.23.0            # Downloads model weights

# --- Document processing ---
unstructured[local-inference]>=0.14.0  # General document processing
pypdf>=4.2.0                          # PDF parsing
PyPDF2>=3.0.0                         # Fallback PDF parsing
PyMuPDF>=1.23.0                       # Better PDF handling for Windows (fitz)
pdf2image>=1.16.3                     # PDF to image conversion
pillow>=10.0.0                        # Image processing

# --- OCR ---
pytesseract>=0.3.10                   # OCR wrapper
# Note: Remember to install Tesseract OCR system-wide

# --- Database ---
# sqlite3 is included with Python, no need to install

# --- Utilities ---
tiktoken>=0.6.0                       # Token counting for LLMs
python-dotenv>=1.0.1                  # Environment variables
werkzeug>=3.0.0                       # File upload handling

# --- Optional but recommended ---
aiofiles>=23.2.1                      # For async file operations (future)
numpy>=1.24.0                         # For embedding operations
scikit-learn>=1.3.0                   # For similarity calculations