AI Trading Intelligence Dashboard

The Challenge

Our client, a supplier of machined aerospace components, operates under strict assembly milestones. Standard Material Requirements Planning (MRP) algorithms forecast demands based solely on static dates, failing to evaluate supplier emails, weather events, or global logistics delays.

This led to unexpected component shortages that stalled assembly lines, forcing the company to maintain a costly 12% safety stock buffer to guard against delays.

The Solution

I designed a predictive inventory dashboard connected directly to Epicor Kinetic databases and the OpenAI API. The workflow includes:

• OData API Pipeline: Querying open BOM lines, pending Purchase Orders, and historical vendor shipping logs.
• AI Analysis: Passing supplier updates and cargo notifications through OpenAI models to identify delay risks and categorize delay likelihood (High, Med, Low).
• Interactive UI Layouts: Building custom dashboard grids inside Kinetic Application Studio that highlight material risks 72 hours before they affect production runs.

The Results

The deployment of the AI dashboard resolved line delays and reduced inventory overhead:

• Proactive Adjustments: Production planners receive material delay warnings 3 days in advance, allowing them to adapt labor assignments.
• Cost Savings: Sourcing confidence allowed the plant to reduce raw safety-stock holding buffers by 12%.
• Forecasting Accuracy: Machine learning forecasts achieved 99.1% reliability over 90 days of live tracking.