Predictive Customer Segmentation for Retail Company

Predictive Customer Segmentation for Retail Company

  • the USA
  • Retail
  • Django
  • Docker Compose + AWS
  • Selenium
  • Tensorflow

The client is a large retailer in the U.S., whose main goal is to create a dynamic retail enterprise that meets customers’ unique needs. The company’s way to gain profits revolves around sending its customers catalogs with clothing brands to the doorstep. No physical shop is present. On viewing the catalogs, customers can order a variety of clothes by mail and phone. For this reason, the target audience of the company is mostly elderly buyers, the average age of whom is 70 years old.

Business needs

When the pandemic started, the idea of selling clothes without physical interaction with clients was considered very profitable and timely. But the business problem of the retail company was that the customer age had to be lowered to ensure the sales would not drop significantly. Thus, the company needed to expand purchases and offer value to younger audiences. As the client didn’t understand what kind of audience they had and what he had to do to attract new buyers, his main goal was to receive an extensive data science consultancy.

inoxoft Solution

Inoxoft has supported the client’s intentions and offered him a hand of help by delivering an ML model to perform categorization. Every word was given an IAB category in the model that could analyze websites and predict consumers’ preferences further. The model was developed as a multilingual classifier giving the possibility to keyword websites written in any other language.

  • IAB categorization.
  • Custom Python Scraping Backend.
  • Custom Tensorflow training and model deployment.

imageProject duration
2020 - Ongoing
imageTeam Composition
  • 1 — Project Manager
  • 1 — Data Scientist
  • 1 — Machine Learning Lead
  • 1 — Data Sofrware Engineers
  • Tensorflow
  • AWS
  • PestoDB/Athena
  • PyTorch
  • Scikitlearn
  • Sagemaker

Predictive Customer Segmentation for Retail Company
Predictive Customer Segmentation for Retail Company
Key Features
  • Utilized AWS Athena + S3 to ingest customer data.
  • Created Tensorflow Deep Neural Networks for Training.
  • Used Sagemaker and Athena to produce on-demand predictions.

Among the challenges Inoxoft engineers have met and successfully overcame was to engage in a lot of DevOps and Data Quality challenges to deliver a successful resut to the client.


Inoxoft’s team of forward-thinking Big Data and Machine Learning experts produced an on-demand data science consultancy and prediction models, which made it possible for the client to:

  • Approach customer data easier and from one place rather than numerous sources.
  • Use modeling and predicting the value on its basis.
  • Understand personal information and purchasing behaviors of 15 million potential customers.
  • Make better predictions according to customer timing, discount levels, brand, and item interest.
  • Obtain models to improve marketing and recommendation systems.
  • Receive 24/7 support from our big data and machine learning experts.
Thank You!

Thank you

Related services