Recommendation Engine for Mattress Selection

Project Summary: Built a machine learning recommendation engine for helping people select their most suitable mattress. I was leading the team responsible for developing the machine learning part of the solution.

Project Duration: 3 months

Team Size: 10 team members, 2 data scientists

Industry: Retail

Customer

Major US retail chain.

Business Case

Sleep is a very important part of everybody’s life and poor sleep because of a bad mattress can lead to unnecessary worsening of a person’s quality of life. However, purchasing a mattress is a difficult and unpleasant experience for most people. Usually, people don’t know what they should look for in a mattress. There is no one-size-fits-all mattress and purchasers need to find the one mattress that suits them best based on their conditions, body shape, lifestyle, etc.

The retail company has a rich assortment of mattresses, having diverse properties, such as, comfort level, support level, price, movement separation, etc. They wanted to boost the customer experience for their online and in-store customers by providing an automated solution for personalized mattress recommendation.

Solution

Our solution was an intelligent recommendation engine that would select the most appropriate mattress for the purchaser. It was built using data from previous purchases of clients as well as their satisfaction with the products that they purchased.

The UI of the solution was a quiz that the clients could fill online or on tablets inside the physical stores. After filling the quiz, the solution would enrich the quiz responses of the customers with additional data that has been previously collected about them. The result is a selection of the 3 best suiting mattresses that the customer can purchase online or try in the store.

Tools & Technologies

  • Google Cloud Platform
  • Google AI Platform
  • Python

My Role & Responsibilities

My role was to lead the development of the machine learning part of the solution. My responsibilities included:

  • Technical consultancy during the discovery phase and design thinking workshops
  • Data science team management
  • Machine learning research & development
  • Defining technical requirements and work estimations
  • Defining the cloud architecture of the ML solution
  • Data analysis
  • Infrastructure setup and deployment of the ML engine