From Diesel to Data: Logistics Prediction and Text-to-SQL Pipelines with AWS AI

In physical commodity trading, manual scheduling of fuel shipments creates operational delays and costly demurrage fees. To streamline operations, Stormit developed two serverless solutions:

• Logistics Prediction Engine: Forecasts shipment timing and parcel quantities, reducing nomination timing error to 0.98 days and saving millions in late fees.
• Conversational Query Assistant: Allows operators to ask plain-English questions to query historical trade data and receive instant answer reports.

Results:

• Cost Savings: Reducing nomination prediction error from 2 days down to 0.98 days eliminates major demurrage fees for idle vessels.
• Infrastructure: Powered by a serverless AWS architecture, removing server management overhead.