Sageglass

Predictive Data, Immediate Impact

Snapshot

For SageGlass, an architectural glass manufacturer and subsidiary of the $45 billion dollar multinational corporation Saint-Gobain, glass is customized to reduce costs for building owners, increase energy efficiency and improve occupant comfort through electrochromic tinting. Our data science work increased yield by 85 percent and provided SageGlass with smarter, more actionable data to make better business decisions.

Contributions

Application Development
Data Strategy
Interaction Design
Data Engineering
Digital Product Delivery

Sage Glass Snapshot

Overview

Business Challenge

Due to the highly-customized nature of SageGlass’ product, waste scrap is unusable. The SageGlass team approached Nerdery wanting to leverage data science to improve the production planning process, reduce overrun and increase the on-time/in-full (OTIF) rate. With a complex challenge and a quick deadline — two months — we quickly ramped up to understand their business, determine targets for OTIF rate and overrun, and define a solution that incorporated a fully-featured, custom-built prediction engine.

Strategy & Solution

Nerdery performed a thorough analysis of their data to understand how it related to the problem at hand. The data was used to build and validate several machine learning models. Those models were integrated into a user interface the SageGlass team could use immediately, providing opportunities for fast business impact. With data science driving real-time production forecasts, profitability was increased by accurately predicting yields, reducing waste scrap and gaining actionable insight into OTIF rate.

Results

02 SageGlass 01 85 %

Yield increase

02 SageGlass 01 copy

Increased profitability