Who is the Client

A US-based manufacturer of semiconductor processing equipment used in silicon wafer processing, with an annual sale of approx. $10B.

The Challenge

The client owns multiple large manufacturing plants in various locations. They have separate units for performing the different processes for manufacturing semiconductor processing equipment. The client had the historical data for surged price (extra fees) that was paid to the suppliers for expediting delivery of semiconductor parts and components but didn’t have a trained model to identify and analyze the factors for predicting the behavior.

The client wanted to identify the hidden surged price pattern and get new insights from predictive analytics.

The Solution

GSPANN identified and analyzed the attributes impacting the surged price. We addressed each suppliers’ behavior while charging extra fees and its association with the delivery time that was affecting the cost. This resulted in the development of a structured and trained model to uncover the hidden price patterns for estimating the surged price.

GSPANN helped the client by solving business problems through a combined power of data science and machine learning techniques. We analyzed the behavior of each supplier as their price surge was incrementing year-over-year (YOY).

Moreover, we trained the predictive model (developed in Azure ML Studio) on historical data of purchases for the last three years, which consists of purchase order data, supplier features, historical transactions, product features, etc. This real-time estimation helped the client in reducing the expedited delivery expenses.

The client was interested in understanding the supplier’s behavior for predicting and estimating the surge in price. To get predictive analysis in a simplified way, we developed the web application’s frontend in Python and Microsoft Azure ML Studio. We extracted the supplier’s historical data from the database through Azure ML API and hosted the ML-trained model on Azure Cloud for sharing the analytical data.

Business Impact

  • The predictive model enabled the client to get the products at a minimum price as per their requirement of raw materials, like semiconductor components and devices.
  • The predictive model was 80% accurate in predicting and estimating the expedited fees before the final order is initiated by the supplier.
  • Improved data governance (including collection and management of data) enabled in keeping the system in place.

Technologies Used

Atlassian Jira and Confluence. Platform to manage sprint stories and provides an online team collaboration environment
Azure Machine Learning (ML) Studio. A GUI-based Integrated Development Environment (IDE) for constructing and operationalizing machine learning workflow
Microsoft Azure. A cloud computing platform for building, testing, deploying, and managing applications and services
Azure Machine Learning (ML) Web Service. A REST API-based model to send data and get predictions that creates and deploys predictive models as web services
Python. An interpreted, high-level, and general-purpose programming language that enables programmers to write clear and logical code
Python ML Libraries. Open-source machine learning libraries

Related Capabilities

Utilize Actionable Insights from Multiple Data Hubs to Gain More Customers and Boost Sales

Unlock the power of the data insights buried deep within your diverse systems across the organization. We empower businesses to effectively collect, beautifully visualize, critically analyze, and intelligently interpret data to support organizational goals. Our team ensures good returns on big data technology investment with effective use of the latest data and analytics tools.

Do you have a similar project in mind?

Enter your email address to start the conversation