28. 📦 Inventory Anomaly Detector

Overview

A time-series analytics app that monitors inventory levels and flags anomalies such as stockouts, overstock, or theft signals.

Primary Use Cases

• Warehouse managers ensuring stock health
• Retail analysts spotting shrinkage or demand spikes
• Supply chain teams optimizing reorder points

Key Features

• Dashboard with real-time stock trends
• Anomaly detection models (ARIMA-based, isolation forest)
• Alerting via email or Slack
• Root-cause analysis suggestions

Tech Stack

• Frontend: React + TypeScript + Recharts
• Backend: FastAPI (Python) + Go for real-time streaming
• AI Models: prophet for baseline forecasting; sklearn.IsolationForest for anomaly detection
• Data: PostgreSQL + Kafka for ingest

Architecture

1. Data Ingestor: streams inventory updates into Kafka
2. Storage: persists to Postgres