E Fundamentals: outpacing e-commerce with Google Cloud Platform
The online retail landscape is growing fast and in constant flux. Built on a flexible, nimble platform that keeps the company in step with industry changes, UK startup E Fundamentals has set itself apart among retail data analysis providers. Rather than offer raw data sets to experts, E Fundamentals creates easy-to-use dashboards that provide actionable insights on e-commerce data.
“Like any startup, we wanted the flexibility to try things out while minimising the costs of mistakes. If we’d started our company 10 years ago, we would have spent hundreds of thousands on hardware, but Google Cloud Platform lets us experiment and deliver without waste. We’ve already moved our applications and data several times, and every week we discuss how new Google offerings might let us do things differently. Keeping pace with Google’s development of APIs, Stackdriver, and Kubernetes Engine makes us feel we’re on the cusp of the wave, always engaging with the latest, greatest tech.” - Adrian Butter, CTO, E Fundamentals
Scaling from scratch
Online retailers are growing and changing everyday, with ever expanding product ranges that retail data analysts need to catalogue. Currently collecting over one million records a week, E Fundamentals runs a daily gather to track products on 20+ major retailers, and data volumes are growing by factors of 2 or 3 annually. From the beginning, the E Fundamentals team knew they needed infrastructure that could scale fast, keep the company ahead of retail development and be cost-effective from the first days of the startup to the global scale.
That’s why E Fundamentals built their client-facing and backend data management systems on Google Cloud Platform. Google Compute Engine eased the company’s transition from a single prototype VM in 2014 up to more than 50 instances today, thanks to Compute Engine's precise pricing and its ability to create and destroy VMs at speed. As the company moves onto the international stage, autoscaling VMs lets engineers manage scaling in minutes, not days, while Google’s server distribution delivers global infrastructure that the small team can run from a single location. In addition, up to 1PB of cost-effective archive on Google Cloud Storage and the inherent stability of Compute Engine's VM system protect the company from catastrophic hardware failures.
“When you start a company you want to play with the all key parts in your architecture to see what works best. Google understands that part, but they also understand that eventually you want tools managed for you. Take Cloud SQL. Rather than set up and maintain your own database, with Google’s SQL database everything under the hood already works and is ready to go. It’s already backed up and running the latest version of the language.” - Adrian Butter, CTO, E Fundamentals
Switching up a gear with Kubernetes Engine
Increasing the depth and detail of the data gathered on products is core to the future success of E Fundamentals, which is why the company now aims to widen its net to include obscure or unwanted product placements online. With Kubernetes Engine, E Fundamentals can flex at high speed between zero and over 100 machines every day, saving time without significant cost increases compared to Compute Engine. Now that it’s established a stable, scalable model, the company has a blueprint to expand into markets worldwide. Additionally, Google Cloud Machine Learning Services are set to play a key role in this expansion, and shifting emphasis from Compute Engine to Kubernetes Engine gives the team space to explore these new possibilities.
“Google is giving every indication of wanting to develop Kubernetes and to make Google Kubernetes Engine the dominant player in its space. Docker containers are at the heart of everything we do, and Google is really shining in this area.” - Adrian Butter, CTO, E Fundamentals