FLAIR – Factory Line Analytics for Insight & Resilience

FLAIR is a data science initiative aimed at analysing KP Manufacturing's production line data to identify inefficiencies, track operational waste, and improve demand forecasting accuracy.

Key Components

• EDA & BI: Analyse shipment and production trends across SKUs and factories.

Folder Structure

• data/: Medallion layers (raw -> bronze -> silver -> gold)
• src/: Modular code (cleaning, modeling, diagnostics)
• notebooks/: Exploratory and structured notebooks for each task

Todos

• Waste Tracking: Reconstruct production flow and estimate yield loss stage-by-stage.
• Forecasting: Predict SKU-level demand using AutoGluon-based time series modeling.