Our website uses cookies

Cookies enable us to provide the best experience possible and help us understand how visitors use our website. By browsing Essential Retail Magazine, you agree to our use of cookies.

Okay, I understand Learn more

Q&A: How data science works at Walmart

Alessandro Magnani, distinguished data scientist at Walmart Labs talks about how data is enabling the business.

Tell us a little bit about yourself and how you became distinguished data scientist at Walmart Labs?
I came to Walmart through the acquisition of my previous company: Adchemy. Adchemy was a start-up in the online ad space. I now work in the catalogue and search teams at WalmartLabs.

I work on product classification and attribute extraction. I also work on projects related to image quality and more recently I have been working on image search and have also been involved in creating a common platform within the company to train and serve machine learning models at scale.

What is the best part of your job?
Working on very interesting problems, access to large data sets and very smart colleagues

What technologies do you use regularly to do your job?
We extensively use TensorFlow, Spark, Hive, and Python. Other teams use many more tools as well.

Tell us about Walmart’s use of data? Has the company always been a big advocate of using data to better the business?
Data has always played a central role in the Walmart decision making process and is one of the most valuable assets the company has. Every decision at all levels is driven by a careful analysis of the data available. For example, the data that is collected daily from each store allows Walmart to understand users’ interest and behaviour at the national and at a local level.

It’s therefore possible to forecast demand and capture trends early on. As another example, the search team relies on accurate product data and analysis of user preference to provide users with the products that they are really looking for.

How is Walmart using data differently?
Walmart collects data both in the physical retail stores as well as online. One interesting challenge is how to use and share this data across these two domains. This requires careful analysis to make sure that the learning from one domain can be used in the other.

How can a business the size of Walmart be agile in its decision making?
A modern data infrastructure is key to process and analyse the enormous amount of data collected. Walmart has invested heavily in big data from the very beginning of its history, but it has invested equally heavily in building an agile, empowered culture. Therefore 'scale'– and specifically economies of scale – is truly always in our advantage.

What worries you about the use of data going forward?
Security of user data is a primary concern. Moreover, we need to make sure that our ability to process and analyse data keeps up with the increase in the amount of data, and need to make decision even more quickly. Finally, it is less about ‘worrying’ and more about ‘we better realise’ the enormous competitive advantage that our big, proprietary data can drive in our business.

Alessandro will be speaking further on his work with machine learning at Argos at the RE•WORK Deep Learning for Retail summit in London next week. Find out more here.

What’s Hot on Essential Retail?