Wayfair expands Google Cloud deal for hybrid cloud strategy

Ahead of NRF, Google Cloud announced a win with Wayfair to expand its hybrid cloud efforts.

How Wayfair approaches natural language processing, machine learning and returns ZDNet's Larry Dignan caught up with Dan Wulin, head of data science and machine learning at Wayfair to talk about how the company is using natural language processing for unstructured data and tracking returns on machine learning investment.

Wayfair is ramping its hybrid cloud strategy with a bet on Google Cloud Platform.

Special report

Data, AI, IoT: The Future of Retail (free PDF)

This ebook looks at the rise of e-commerce and the digital transformation of retail companies.

Read More

The win for Google Cloud is significant since it highlights inroads to the retail and e-commerce sectors. Both Google and Microsoft Azure have pitched their clouds to retailers worried about competition from Amazon, owner of Amazon Web Services.

Google Cloud announced the Wayfair deal ahead of the National Retail Federation's annual conference in New York. 

Wayfair started with Google Cloud for burst capacity and expanded to compute, storage, networking, analytics, and G Suite. The furniture retailer is looking to scale and leverage Google's machine learning and analytics know-how. Wayfair already has an in-house data analytics unit but is using  Google Cloud to help generate insights. 

Must read:

Steve Crusenberry, vice president of infrastructure and platform engineering at Wayfair, walked ZDNet through the Google Cloud decision.

How did you break down the Google Cloud decision vs. AWS and Azure? Did Amazon as a competitor matter?

Google's cloud services are an ideal match for our hybrid cloud infrastructure. The ease at which we could connect our automation platform and our similar SRE-based (Site Reliability Engineering) operating models were selling factors. Additionally, we have a long-standing partnership with Google overall, so it made sense to extend that partnership into Google Cloud.   

We've spoken with Wayfair before about its machine learning and AI efforts. What did Google bring to the table on that front?

We have a team of more than 2,300 engineers and data scientists that build our machine learning algorithms in-house. As we've continued to grow to a massive scale, Google Cloud is helping us store and process the vast amounts of operational data that we use to drive powerful, seamless experiences across merchandising and personalization, as well as marketing and promotion campaigns.

You're going hybrid, so what's the optimal ratio between cloud and your own infrastructure in the next five years. What's the optimal ratio?

Our hybrid cloud strategy is to completely blur the lines between our data centers and the cloud. We move workloads dynamically between our data centers and the Google Cloud to meet peak capacity and processing demands of our website and related applications. We use real-time performance and customer experience metrics to help us determine the correct ratios to use throughout the day.