How ralo93/customeranalyticsdashboard works
Project type and purpose
• This is a full-stack, microservices-based data processing and visualization platform (“Customer Conversation Analytics”).
• It ingests raw text snippets, extracts features, classifies them via AI, stores all results, caches for quick access, and presents an interactive dashboard.
File structure
• Top-level folders and files:
• data/ – raw JSON knowledgebase files
• images/ – static assets
• main.py – entry script (likely ties into honcho Procfile)
• docker-compose.yml & Procfile – orchestration configs
• ml_insights_system/ – root for all microservices
◦ api_gateway/ – Nginx + Python REST API
◦ dashboard_service/ – Streamlit dashboard
◦ db_service/ – handles PostgreSQL operations & models
◦ feature_service/ – extracts features from sentences
◦ queue_service/ – abstracts RabbitMQ enqueue logic
◦ worker_service/ – consumes queue, calls OpenAI, writes back
• tests/ – EDA scripts, ingestion scripts, unit tests for each service
README for additional insights
• Architecture described as flowchart: User → API Gateway → Feature Extractor + Queue Service + DB Service → RabbitMQ → Worker → OpenAI → DB → Redis → Dashboard
• Tech stack: Python 3.12, Nginx, Poetry, Streamlit, OpenAI API, PostgreSQL, RabbitMQ, RESTful API, Docker
• Emphasis on microservices, fast iterations, cheap scalable MVP