IntelliInspect
Real-Time Predictive Quality Control - Software Architecture
Frontend Tier
Backend Tier
ML Tier
Data Layer
Docker Container
USER
FRONTEND
BACKEND
ML SERVICE
DATA LAYER
👤
User
🐳 Docker Container
Nginx
Proxy Server
:4200
Angular
Frontend UI
v18+
Static
Assets
🐳 Docker Container
ASP.NET Core API
Orchestration Layer
Auth • Proxy • JWT
:5000
🐳 Docker Container
FastAPI ML Service
Python 3.13 • XGBoost
Training • Simulation
:8000
🐳 Container
Redis
Cache
File System
Docker Volumes
data_store/
models/
Parquet • XGBoost
HTTP/HTTPS
Assets
API Calls
/api/*
REST API
WebSocket
Session Data
Persist Models
API Endpoints:
/api/Auth/login
/api/Data/upload-chunk
/api/Data/finish-upload
/api/Model/train
/api/Model/simulation-ws
ML Endpoints:
/csv/upload-chunk
/csv/finish-upload
/csv/validate-ranges
/csv/train
/csv/simulation-ws
🔗 Docker Network: abb-network
Workflow Steps:
1. Upload Dataset (Chunked)
2. Configure Date Ranges
3. Train XGBoost Model
4. Real-Time Simulation
Technology Stack:
• Angular v18+
• ASP.NET Core 8
• Python 3.13 + FastAPI
• Redis + File Storage
• Docker + Compose
Data Formats:
• CSV Upload (2.5GB max)
• Parquet Processing
• XGBoost Models
• Real-time JSON streams