Machine Learning Saves Costs and Improves Pricing Accuracy

Gaining efficiency through data-driven automation

Challenge

Lower costs and increase product classification accuracy

This global heavy equipment manufacturer prices its parts based on what class they belong to. To lower staff cost and increase accuracy, the pricing division sought to automate how parts  get classified, and by extension, how they are priced. 


Solution

Leverage machine learning to automate & improve processes

Clarity’s data scientists leveraged machine learning techniques and a variety of algorithms to show that the process of classifying parts can be successfully automated. The primary challenge was the slow run-time associated with the computational scale of the problem (large feature set over millions of
data points). 

Clarity tackled the data issue by applying a variety of dimensionality reduction techniques and solved the computational problem utilizing Apache Spark™ in a distributed computing environment. 

Outcome

Saved on workforce costs, streamlined classification process, improved pricing

The client currently utilizes a staff of over 50 employees to classify and price new parts.  The potential automation of this process will allow the client to significantly reduce workforce costs, increase efficiency and more accurately price new parts impacting profit margins and customer goodwill.

Clarity’s team was able to identify and leverage the most influential features to classify new parts with a high degree of accuracy. Additionally, our algorithm was able to identify parts that were previously misclassified by human analysts. 

Contact us

Technologies

apache spark

Latest News Header

Latest News Subheader

Recognizing Women at Clarity

Happy International Women’s Day to all our women in Tech and personal Shero’s! We want to take advantage of this opportunity to recognize all the strong women we get to work alongside everyday at

The Data Model in The Era of Snowflake Computing’s Data Warehouse: SMP and MPP

Three-Part Series: Article 3 In the first article of this series I questioned whether the continued comparison of Relational Models to Star Schema models is still relevant with the entrance of

The Data Model in The Era of Snowflake Computing’s Data Warehouse: Business Architecture and Star Schema

Three-Part Series: Article 2 In the prior article I questioned whether the continued comparison of Relational models to Star Schema models is still relevant with the advent of Snowflake Computing’s

why-clarity.jpg

Find out how Clarity Insights can help you 

Contact us