requirements.txt

# Real-Time Document Intelligence & Compliance Engine
# Python Dependencies
# Version: 1.0
# Date: 2026-02-03

# =============================================================================
# CORE FRAMEWORK DEPENDENCIES
# =============================================================================
# --extra-index-url https://download.pytorch.org/whl/cpu
# Web Framework
fastapi==0.104.1
uvicorn[standard]==0.24.0
pydantic==2.5.0
pydantic-settings==2.1.0

# Database Dependencies
psycopg[binary]==3.1.13
asyncpg==0.29.0
sqlalchemy==2.0.23
alembic==1.13.1

# Redis and Caching
redis[hiredis]==5.0.1
aioredis==2.0.1

# Neo4j
neo4j==5.15.0
py2neo==2021.2.4

# =============================================================================
# TASK QUEUE AND ASYNC PROCESSING
# =============================================================================

# Celery and Task Processing
celery[redis]==5.3.4
flower==2.0.1
kombu==5.3.4

# Async Support
asyncio-mqtt==0.16.1
aiofiles==23.2.1
aiohttp==3.9.1

# =============================================================================
# STORAGE AND FILE HANDLING
# =============================================================================

# Object Storage (S3/MinIO)
boto3==1.34.0
minio==7.2.0

# File Processing
python-multipart==0.0.6
python-magic==0.4.27
Pillow==10.1.0
PyPDF2==3.0.1
pdf2image==1.17.0
python-docx==1.1.0

# =============================================================================
# MACHINE LEARNING AND NLP
# =============================================================================

# Deep Learning Framework
torch==2.1.2+cpu
torchvision==0.16.2+cpu

# Transformers and NLP
transformers[torch]==4.36.2
sentence-transformers==2.2.2
tokenizers==0.15.0
datasets==2.16.1

# Hugging Face Hub
huggingface-hub==0.20.1

# Traditional NLP
spacy==3.7.2
nltk==3.8.1
textblob==0.17.1

# =============================================================================
# COMPUTER VISION AND OCR
# =============================================================================

# OCR Libraries
pytesseract==0.3.10
easyocr==1.7.0

# Image Processing
opencv-python-headless==4.8.1.78
scikit-image==0.22.0

# Layout Analysis
layoutparser[ocr]==0.3.4
detectron2 @ git+https://github.com/facebookresearch/detectron2.git

# =============================================================================
# VECTOR DATABASES
# =============================================================================

# Pinecone
pinecone-client==2.2.4

# Weaviate
weaviate-client==3.25.3

# Alternative Vector Stores
chromadb==0.4.18
qdrant-client==1.7.0

# =============================================================================
# DATA SCIENCE AND ANALYTICS
# =============================================================================

# Scientific Computing
numpy==1.26.2
pandas==2.1.4
scipy==1.11.4

# Machine Learning
scikit-learn==1.3.2
xgboost==2.0.2
lightgbm==4.1.0

# Anomaly Detection
pyod==1.1.3
isolation-forest==0.0.1

# Statistical Analysis
statsmodels==0.14.1

# =============================================================================
# SECURITY AND AUTHENTICATION
# =============================================================================

# JWT and OAuth
pyjwt[crypto]==2.8.0
python-jose[cryptography]==3.3.0
passlib[bcrypt]==1.7.4

# Cryptography
cryptography==41.0.8
bcrypt==4.1.2

# OAuth2
authlib==1.3.0

# =============================================================================
# MONITORING AND OBSERVABILITY
# =============================================================================

# Metrics and Monitoring
prometheus-client==0.19.0
opentelemetry-api==1.21.0
opentelemetry-sdk==1.21.0
opentelemetry-instrumentation-fastapi==0.42b0
opentelemetry-instrumentation-sqlalchemy==0.42b0
opentelemetry-exporter-prometheus==1.12.0rc1

# Logging
structlog==23.2.0
python-json-logger==2.0.7

# Health Checks
healthcheck==1.3.3

# =============================================================================
# VALIDATION AND SERIALIZATION
# =============================================================================

# Data Validation
marshmallow==3.20.2
cerberus==1.3.5
jsonschema==4.20.0

# Date and Time
python-dateutil==2.8.2
pytz==2023.3

# =============================================================================
# CONFIGURATION AND ENVIRONMENT
# =============================================================================

# Configuration Management
python-dotenv==1.0.0
dynaconf==3.2.4

# CLI Tools
click==8.1.7
typer==0.9.0

# =============================================================================
# TESTING DEPENDENCIES
# =============================================================================

# Testing Framework
pytest==7.4.3
pytest-asyncio==0.21.1
pytest-cov==4.1.0
pytest-mock==3.12.0
pytest-xdist==3.5.0

# Test Fixtures and Factories
factory-boy==3.3.0
faker==20.1.0

# HTTP Testing
httpx==0.25.2
requests-mock==1.11.0

# Database Testing
pytest-postgresql==5.0.0
pytest-redis==3.0.2

# =============================================================================
# DEVELOPMENT TOOLS
# =============================================================================

# Code Quality
black==23.12.0
isort==5.13.2
flake8==6.1.0
mypy==1.8.0
pylint==3.0.3

# Pre-commit Hooks
pre-commit==3.6.0

# Documentation
mkdocs==1.5.3
mkdocs-material==9.5.3
sphinx==7.2.6

# =============================================================================
# UTILITY LIBRARIES
# =============================================================================

# HTTP and Networking
requests==2.31.0
urllib3==2.1.0

# JSON and Data Formats
orjson==3.9.10
pyyaml==6.0.1
toml==0.10.2

# Text Processing
regex==2023.10.3
unidecode==1.3.7
langdetect==1.0.9
googletrans==4.0.0rc1

# File System and OS
pathlib2==2.3.7
watchdog==3.0.0
psutil==5.9.6

# Concurrency
concurrent-futures==3.1.1
threading2==0.1.2

# =============================================================================
# EXTERNAL API INTEGRATIONS
# =============================================================================

# AWS Services
aioboto3==12.0.0

# Cloud Services
google-cloud-documentai==2.20.1
azure-cognitiveservices-vision-computervision==0.9.0

# Business Integrations
salesforce-api==0.1.85
python-docusign-esign==1.0.0

# Communication
slack-sdk==3.26.1
twilio==8.11.0

# =============================================================================
# PERFORMANCE OPTIMIZATION
# =============================================================================

# Caching
cachetools==5.3.2
diskcache==5.6.3

# Memory Management
memory-profiler==0.61.0
pympler==0.9

# Parallel Processing
joblib==1.3.2
multiprocessing-logging==0.3.4

# =============================================================================
# DEPLOYMENT AND INFRASTRUCTURE
# =============================================================================

# WSGI/ASGI Servers
gunicorn==21.2.0
uvloop==0.19.0

# Docker and Containerization
docker==6.1.3

# Configuration for different environments
python-decouple==3.8

# =============================================================================
# VERSION CONSTRAINTS
# =============================================================================

# Ensure compatibility
setuptools>=65.0.0
wheel>=0.38.0
pip>=23.0.0