AI Market Price Prediction in Transportation and Logistics

With a fleet of over tens of thousands of trailers and containers that performs approximately million annual deliveries, our client is one of the Fortune 500 transportation and logistics company. To stay ahead of competitors, they partnered with Infosys to predict accurate market price, using AI and ML solution, that helps them arrive at the best carrier cost and shipper price resulting in optimal margin.

The Objective

Develop an AI solution to predict market price with accuracy > 90% that will help negotiate better in spot market broking business
Increase the bid-win rate from the current 1%
Better procurement cost with the potential carriers by providing accurate carrier cost for every spot load, considering lane and load attributes
Better shipper price for shipments to increase the gross margin per employee

Key Challenges

Client deals in spot market and often face problems such as over-paying carriers or losing business from shippers due to inaccurate prediction of their market price.
This challenge was mainly due to high error in their existing market price prediction model which affects their bid-win rate and margin significantly
Hence there is a need to predict optimal market price that pays carrier the right cost and win the load from shipper with optimal margin

Ready to experience?

TALK TO EXPERTS

The Solution

Infosys Solution

Infosys delivered a novel dynamic pricing approach by predicting market price that gives maximum margin using cognition and hyper-automation. The dynamic pricing is powered by advanced ensemble-based machine learning approach with multiple models implemented across lanes considering relevant features. The power of Artificial Intelligence is leveraged to reduce the manual effort and time consumed to determine the market price. The solution is hosted in cloud environment and the entire pipeline is automated starting from extracting data from legacy system, data quality, transformation and loading, feature engineering, modeling, and integrating with the consuming systems.