Get in touch with us. We'd love to hear from you.
Crowdsourcing Data Science Solutions for Retail from Across the Globe
When you think you’ve found the right technology partner to help develop and deploy solutions in APAC, it’s vital that you ask, “What sets this vendor apart from all of the other vendors serving the region?” There are many specific elements to that answer depending on the domain, technology and strategy aspects of a project, but one recurring theme is clear: A wealth of global experience and a network of experts is invaluable when it comes to adapting to the region’s challenges, no matter how unique or specific they are.
A great example of how valuable this network can prove to be happened recently when one of the largest retailers (and an existing EPAM client) here in Hong Kong posed a complex data science challenge. In a matter of days we were able to mobilize our global community of data scientists and developers and leverage our DLab rapid analytics toolset to get started on the problem, and after a couple of weeks we were able to demonstrate the requested results using data provided by the retailer.
This client’s specific challenges focused on a range of retail needs, like being able to recommend products and upsell based on previous consumer engagement and transaction data. We vetted the solution by using part of the existing historical data to check if our Recommender provided promotional suggestions that matched with the actual purchases of the consumers.
Based on their findings, our data scientists proposed a Hybrid Recommender model-based approach, which helps provide personalized recommendations not only for consumers with an extensive purchase history, but also for new or new-ish customers who can get semi-personalized recommendations based on context. Here’s how it works:
Another thing to look for in a technology partner is the ability to take a holistic approach to a project to keep things moving along. As an example, at the same time as we were conducting research on our retail customer’s data science challenge, we had experienced consultants landing in Hong Kong to get the work started. Furthermore, since we were able to demonstrate the business value of the solution and capture the interest of one client, immediately other local opportunities emerged, and we’re starting to recruit and build a local competency in this area. We were proactive, quick to react, and now we have a new differentiated capability in the region.
Beyond enabling personalized recommendations, retail-related data science capabilities can be leveraged for other customers in APAC. For example, we have developed a solution for analyzing the composition and availability of items on store shelves. Employing the latest advances in deep learning, we are able to automatically recognize from in-store images not only individual products, but also the type and size of the package. This helps our client to improve operations and automate repetitive manual tasks.
To connect the dots to another vertical, artificial intelligence and machine learning/machine vision are also leveraged in financial services, not only for similar recommendation and robo-advisory systems, but also for a lot of robotics and automation use cases. I’ll share some recent examples in my next blog – stay tuned!