PROPOSED AI APPLICATION

Metal order demand forecaster

Material buying misses customer demand

Steel and metal processing businessesHistorical data model pilotProposed solution scope

The problem and the proposed outcome

Application: Steel and metal processing businesses — Metal order demand forecaster
Problem to solve: Material buying misses customer demand
Target buyer: Metal processors, stockyard owners
Intended use: Employee / expert own workflow; internal business solution
Benefit to test: Test whether numeric estimates improve planning outcomes under realistic uncertainty.
Potential delivery: Internal application, reviewer dashboard, private assistant or authorized user portal; choose one surface after validating the workflow.

AI method: Forecasting and numeric prediction
Method code: F
AI contribution: Learn relationships from historical targets and relevant context to estimate a future quantity or numeric outcome; present uncertainty where supported.
Output artifact: Quantity, duration or cost estimates with context and uncertainty notes.
Core modules: Historical data intake → baseline comparison → prediction → planner review
Simple baseline: Recent averages, seasonal naive forecasts, median effort and spreadsheet scenarios.
Why test AI: Could help when multiple contextual factors affect demand, duration or costs; improvement over a simple baseline must be demonstrated.

Pilot track: Historical data model pilot
Required data: Orders; consumption history; lead times
Data readiness: Enough representative history at the required granularity, known target definitions and leakage-free features are needed; quantity adequacy is task dependent.
First development step: Define the target and decision horizon, inspect historical completeness and measure a simple baseline before modelling.
Initial MVP scope: One defined task, one permitted dataset and one reviewer workflow. Manual upload or authorized export first; no unattended external action. Validate the model before live integration.
Demo data: Synthetic or licensed public samples may illustrate the workflow. Real benefit and model performance need representative authorized held-out data.
Buyer validation question: Ask Metal processors, stockyard owners: show a recent case where material buying misses customer demand, record the present effort or error, and confirm whether the required data can lawfully be used.

Primary success metric: Stockout rate
Model evaluation: Use temporal or entity-separated hold-outs as appropriate; compare MAE and relevant business error against naive, seasonal or median baselines.
Metric target: Set after baseline measurement; no improvement or accuracy result asserted.

Specialist review: No special flag; domain review still applies
Human review boundary: Authorized planner reviews assumptions and constraints before changing any plan, procurement or resource allocation.
Domain review: Qualified technical teams approve specifications and procedures. The first MVP provides reviewed evidence or planning; it does not directly control machines or safety systems.
Related micro-ERP catalogue IDs: IWA-B-0838D60FEAD0; IWA-B-30269EC6920A; IWA-B-081AAD7F76D3
Integration status: Potential segment link only; no product integration implemented.