Using Convect and Zapier to Deploy Machine Learning Solutions for E-Commerce.

submited by

Style Pass

2021-06-26 01:30:04

A machine learning (ML) algorithm hosted on Convect is going into production this week, estimated to help double manufacturing throughput for a custom garments e-commerce business.

In this post, we’ll describe an operational challenge at this company (hereinafter referred to as X), the solution they’ve implemented with Convect to grow with demand, and how Convect is making it easy for data scientists to deliver no-code solutions to customers. This work exemplifies both (1) a need for data science in direct-to-consumer e-commerce and (2) the value of a workflow that enables data scientists to meet that need without leaving a Jupyter notebook and clients to utilize ML solutions without writing code.

To further collapse time to value for customers, we’re in the process of rolling out an integration with Zapier, a no-code tool for automatically moving data between web apps. You can try out the Convect-Zapier integration by clicking on this link, and a step-by-step tutorial for getting started with the integration can be found in the docs.

X’s current process is challenging to scale. Customers submit measurements from around their body, and X’s pattern maker generates and applies a custom pattern on fabric according to these measurements from which the garment is cut and sewn. X spends about 1.2 hours creating each custom pattern.