How analytica-manulife/aiservices works
Project type and purpose
• From the file tree and README, this is a Python-based AI Services collection (“AIServices”) composed of several analysis and generation scripts (e.g. NewsAnalyzer, KYCAnalyzer, ImageGenerator, StockSuggest, TransectionAnalysis).
• It is a backend CLI/service application for performing different AI/ML–powered analyses, with a central orchestrator (main.py) that invokes specialized modules and writes/reads data from local JSON/CSV stores.
File structure
• Top-level scripts:
• main.py – the entry point that wires everything together
• AnalyzerFunction.py – common utilities or shared model-invocation code
• ImageGenerator.py – module for AI image creation
• NewsAnalyzer.py – module for news sentiment/insight analysis
• KYCAnalyzer.py – module for “Know Your Customer” document analysis
• StockSuggest.py – module to suggest stocks based on data
• TransectionAnalysis.py – module for transaction data analysis
• Data directories/files:
• company_news/*.json – cached news data for various companies
• news.csv – a flat-file log of incoming news items or analysis results
• Configuration:
• requirements.txt – lists Python dependencies (e.g., TensorFlow, PyTorch, scikit-learn, requests)
• .gitignore – standard ignores
• No separate “frontend” or microservice directories indicate a monolithic service-oriented design.
README for additional insights
• README is minimal (“# AIServices”), so all architectural understanding comes from the file names and directory layout.