Yusp

The Customer

Yusp (by Gravity Research & Development) creates recommendation systems that automatically analyze and learn from browsing behavior to help users find new relevant online content.

When asked about the company vision, Peter Szabo, Global Services Manager at Yusp by Gravity R&D, was quick to answer: “Be relevant or be forgotten.” And with over 35 billion personalized recommendations churned out every month, millions in extra revenue for big-name clients, and a tie for first place in The Netflix Prize competition for best collaborative filtering algorithm, he has a point. Yusp is well on its way to cementing their name in the personalization engine hall of fame.

The Yusp recommendation engine, developed by the Gravity R&D team, provides a scalable personalization solution for enterprises across various industries, from retail websites to dating sites. The engine predicts individual behavior and preferences to provide personalized recommendations to create a better shopping experience and help facilitate product and content discovery.

The Challenge

Yusp needed high processing power and a way to instantly ramp up compute capacity while still maintaining reliability and speed to deliver results in near real-time. "We have marketplace customers with up to 20 million different items, video streaming providers with huge traffic peaks of up to 100 million daily active users, and clients who need us to be present in multiple locations,” says Szabo. Because of these challenging requirements, Yusp also required a personalized customer service element if they had any technical or business issues.
 

While traditional dedicated servers could provide the processing power Yusp required, they weren’t easily scalable at times of high demand. In addition, Yusp faced limitations with their previous hosting provider due to a lack of certificates, limited flexibility, and scarce SLAs for sensitive clients.

The Solution

Gabor Vincze, Senior Consultant of Infrastructure and Services at Yusp, led an extensive research process to determine which service provider would be the best long-term partner for the growing company. When he discovered Leaseweb, he felt the click instantly. “In a short time, we became close partners with Leaseweb,” says Vincze, “and considering our infrastructure needs and architecture at that time, our ROI with Leaseweb was significantly better than moving to any other hyperscalers. Leaseweb was able to provide all technical competencies alongside the high performance and flexibility that you can usually only get from hyperscalers.”
 

Happy with the initial dedicated servers Leaseweb provided and facing growing demand, Yusp soon switched up to a hybrid cloud model utilizing dedicated servers and on-demand Leaseweb Elastic Compute services. The highly customized solution enabled Yusp to increase capacity and scale up instantly during demand surges – all at a substantially lower cost compared to large hyperscalers.

The Outcome

“We saw the biggest difference in reliability and flexibility, especially with the hybrid cloud setup we have at Leaseweb,” says Vincze. “Significant scalability allows us to provide up to 35 billion personalized recommendations a month. We’ve continuously met our stringent SLAs with clients for the past seven years.”
 

Just as Yusp’s vision of relevancy has only strengthened throughout the years, so has their relationship with Leaseweb. “From the beginning, our operations grew together to match each other’s needs,” Vincze says. “Ours has and always will be a two-way relationship.”