Handles 350 million requests per day
Processes 100,000 events per second in production
Scales to accommodate seasonal traffic spikes with minimal disruption
Simplifies internal reporting with a data warehouse built with BigQuery
Empowers non-technical staff to test ML models with Vertex AI to accelerate innovation and collaboration
With Google Cloud, Tabby has built a highly scalable, resilient service that enables users to buy now, pay later, and boost financial inclusion.
With Google Cloud, Tabby has built a highly scalable, resilient service that enables users to buy now, pay later, and boost financial inclusion.
"When people think of fintech, it can often seem esoteric or a bit abstract," says Alex Shchepetilnikov, CTO of Tabby. "But when I look at our reviews and see comments about how we helped people pay for a dentist, or their kids' school fees - it's that kind of real-world impact that makes my job really rewarding."
Tabby helps people stretch their money further with a buy now pay later app that costs users nothing and allows vendors to boost sales, secure in the knowledge that they'll be paid.
After twenty years of building companies up, I knew that only Google Cloud could provide the kind of easy to use, reliable service that could get us up and running, with scalability built-in.
Alex Shchepetilnikov
CTO, Tabby
Founded in 2019 by a team of just four, Tabby is now accepted by more than 40,000 vendors across the UAE, Kingdom of Saudi Arabia, and Kuwait. The company has been on a remarkable journey, accompanied by Google Cloud since its very first step.
Building the infrastructure for a fintech startup can be a tricky business. While it's all well and good perfecting the right architecture and design, at an early stage, the perfect risks becoming the enemy of the good. "You have to find the right balance between simplicity, which you need at the start, and scalability for when you take off as a company," says Shchepetilnikov. "After twenty years of building companies up, I knew that only Google Cloud could provide the kind of easy to use, reliable service that could get us up and running, with scalability built-in."
The building blocks of scalability
From the start, Tabby's setup has been built around Google Kubernetes Engine, which combines the flexibility of container technology with the convenience of a managed service, and supported with Pub/Sub, the low-latency messaging and queueing service. Meanwhile, Cloud SQL handles Tabby's database needs.
For Shchepetilnikov, the key to success has been in choosing the right tools for Tabby from the outset. "GKE, Pub/Sub, and Cloud SQL are really the core pillars of our infrastructure, and they've been that way from the start," he says. "What's remarkable is how little we've had to change our architecture. Instead, we just scaled up."
GKE, Pub/Sub, and Cloud SQL are really the core pillars of our infrastructure, and they've been that way from the start. What's remarkable is how little we've had to change our architecture. Instead, we just scaled up.
Alex Shchepetilnikov
CTO, Tabby
That doesn't mean, however, that Tabby has stayed static. While the core infrastructure design may not have changed much, the company has shown a keen interest in expanding into data and AI. Bigtable and BigQuery, for instance, have helped Tabby to gain and share deep, strategic insights from its data. "We built our data warehouse with Bigtable and BigQuery, and a lot of things are built on top of this, from daily dashboards to investor strategy reports," says Shchepetilnikov.
In addition, the data warehouse has led to Tabby's most recent forays into artificial intelligence using Vertex AI, for customer recommendation engines and risk models. Vertex AI enables non-technical users like risk analysts to test the models they need quickly and easily, erasing dependencies on other teams, accelerating innovation, and strengthening collaboration.
"Previously, our risk analysts would write a business requirements brief for the data science team or the engineering team to create a feature or a model to the system," explains Shchepetelnikov. "With Vertex AI, they can play around with models and test data themselves and get some real results, which makes collaborating on a project with engineers much easier."
Expanding the portfolio, furthering the mission
With other infrastructure solutions, during peak periods like Ramadan or Black Friday, you have to set up 24/7 war rooms. Google Cloud makes traffic spikes very easy for our DevOps team. With a little bit of planning, it's not much different from a regular day at work. It means our team can focus on strategy rather than firefighting.
Alex Shchepetilnikov
CTO, Tabby
Tabby's growth has been remarkable. Since its launch in 2019, Tabby has expanded to serve more than 20 million registered shoppers and 40,000 active sellers, driving annualized sales of more than $14 billion across the region. Such rapid growth can lead to stress on a company's IT infrastructure.
With Google Cloud, however, Tabby has built the ideal foundation to handle whatever challenges it encounters. With GKE and Pub/Sub, Tabby's production cluster now handles 100,000 events per second and processes around 350 million requests per day. The growth was so seamless, the team "barely noticed it happening," as Shchepetilnikov puts it.
Working with retail partners, Tabby has to be able to handle sudden spikes in traffic. What sets Google Cloud apart for Shchepetelnikov is how easily it handles them: "With other infrastructure solutions, during peak periods like Ramadan or Black Friday, you have to set up 24/7 war rooms," he says. "Google Cloud makes traffic spikes very easy for our DevOps team. With a little bit of planning, it's not much different from a regular day at work. It means our team can focus on strategy rather than firefighting."
From the start, Tabby's mission has been about giving people control of their finances. While flexible payments remain their flagship offering, the company is now laying the groundwork to move beyond credit with spending accounts and money management tools. Following the launch of the Google Cloud region in Dammam in 2023, Tabby has been hard at work on expanding its portfolio in Saudi Arabia.
"I can't overstate the importance of the Dammam Cloud Region for helping us to build out our financial ecosystem," says Shchepetelnikov. "With our infrastructure physically located in the country, we can show the regulators that we are complying with all their data sovereignty requirements. And I know that with Google Cloud, we can rely on the same building blocks and high level of support no matter how far we scale."