Digital photo-book company Lupa migrated to Google Cloud Lupa to improve infrastructure scalability during seasonal peaks in demand. And by accessing NVIDIA GPUs, it’s also built a foundation for a faster, machine learning-driven future.
Thanks to smartphones, nearly all of us walk around with high-quality cameras in our pockets every day. They enable us to capture an important moment in seconds and help us document everything we see. In fact, according to Statista, over 1.2 trillion digital photos were taken in 2017.
But too often, these photos are lost and forgotten among the thousands that fill our phones’ storage. Israel-based Lupa saw that as an opportunity, and created mobile and desktop apps that enable customers to turn their digital images into physical, high-quality photo-books. And with over 200,000 customers – representing over 50% of the market – it’s proven extremely popular.
As the company rapidly grew, Lupa needed to find a way to scale its infrastructure to effectively meet the rising demand. Previously, Lupa’s team maintained its physical servers in a location outside the city. But without a dedicated development and operations team, it was a time-consuming process that presented a significant barrier to scalability.
Lupa wanted to migrate its infrastructure to a cloud platform that could both reduce the time spent on maintenance, and accommodate peaks in demand during the holiday seasons. It knew that moving to the cloud could offer it the flexibility and reliability required – it just needed to find the right partner.
The power to meet customer demand
Operating on Google Cloud with NVIDIA GPUs, Lupa has the flexibility to add more compute power in seconds whenever it’s needed, making it much easier to scale quickly when demand rises.
The additional compute power has also made a big impact on its service. Now, during holiday seasons, Lupa can accommodate up to 70% more business than it could running on a physical infrastructure – creating a smoother experience for its customers and helping the company complete more sales.
The migration has made maintenance easier for Lupa’s team too. “Previously, I spent a lot of time on networking, making sure routers were up-to-date, and other tasks,” says Shlomi. “Now everything is under one umbrella, which makes it easier to see which areas we can optimize, and also gives us a lot of flexibility.”
With the compute power of NVIDIA GPUs behind them, Lupa also has the flexibility to experiment without affecting its customer-facing service. “It’s not just the ability to add extra capacity that’s important, it’s the ability to experiment,” says Shlomi. “You can build or duplicate an environment very quickly, and there’s no worry that the development environment will affect production. That’s real architectural freedom.”
Now, Lupa is looking to further expand its architecture and continue building on its service. “We’re really excited about growing our architecture further,” says Shlomi, “all thanks to the freedom to experiment we’ve gained from migrating to the cloud and adopting NVIDIA GPUs.”
With machine learning high on its digital agenda, Lupa will continue to lean heavily on both the AI capabilities and platforms delivered in Google Cloud, and the unparalleled AI performance delivered by NVIDIA GPUs.