Grip predicts shipping delivery success to recommend optimal logistics

Overview

Grip helps e-commerce companies ship perishable goods (meal kits, frozen food, flowers, pet food) by processing hundreds of thousands of monthly orders plus shipment tracking, weather and customer feedback data on the Databricks Data + AI Platform to predict the probability of delivery success and dynamically recommend the best carrier, insulation, packaging and shipping logistics. Grip builds machine learning models and data engineering pipelines end-to-end on Databricks; customers using its recommendations see a 25% reduction in damage rates and a 30% reduction in shipping costs.

The challenge

Shipping perishable goods like meal kits, frozen food, pastries, pet food and flowers reliably across the U.S. requires processing many scattered data points, from shipping carrier tracking records to weather forecasts and customer feedback, to make effective shipping recommendations.

The solution

Grip built machine learning models and data engineering pipelines end-to-end on the Databricks Data + AI Platform, using Delta Lake, Lakeflow Jobs and Unity Catalog to consolidate data from sources like Shopify, ShipStation, warehouse systems, weather APIs and carrier pricing, to predict the probability of delivery success and dynamically recommend the best carrier, refrigerant, insulation and packaging.

Reported business value

Customers using Grip's recommendations see a 25% reduction in damage rates and a 30% reduction in shipping costs, while adopting Databricks SQL Serverless decreased dashboard refresh time by 50% and increased exploratory productivity time-to-insights by 50%.