Varo launched in 2017 on a mission to disrupt the slow and technologically-limited banking industry with a human-centric, tech-first bank designed to help their customers improve their financial health.
Just 3 years later, Varo became the first-ever US-consumer fintech to be granted a national bank charter and in the following 13 months, officially launched Varo Bank N.A., doubled its opened accounts to 4 million, and tripled its revenue.1
Designing an all-digital bank that could deliver on Varo’s mission, differentiate in a crowded market, and comply with constantly evolving regulatory demands was no easy task. So when Deep Varma, CTO at Varo, and his team began to build Varo Bank, they knew they needed to start from day one with a technology stack flexible enough for rapid innovation yet secure enough for a bank.
“Really, it’s speed and agility that’s going to take you to the next level, rather than the technology itself. That’s why I get so excited about the supergraph.”
CTO at Varo
“My strategy was always to use a supergraph because of the benefits of its federated schema,” Varma said. “The graph can play an instrumental role in the schema definitions where my data is saved, but more importantly it can change rapidly based on what I need to do.
“People in the market like to talk about using the best technology, but eventually, everyone is going to be using GraphQL. So really it’s speed and agility that’s going to take you to the next level, rather than the technology itself. That’s why I get so excited about the supergraph.”
The supergraph provided Varo’s engineering team with the foundation they needed to efficiently scale their platform. Sundar Siva, Sr. Director of Engineering at Varo explains that their decision to leverage a federated graph cut years off of their initial time to market:
“Before we launched Varo Bank, we had Varo Money, a payments app which utilized the standard API gateway in a monolithic approach. But given the number of microservices we needed to build Varo Bank, using a monolith would have set us back years. We didn’t have that kind of time,” says Siva.
“With Varo Bank, we had the opportunity to build a brand new tech stack from scratch. We knew we had to go with GraphQL and Apollo Federation. There’s no other viable solution for composable architectures at scale.”
“We knew we had to go with GraphQL and Apollo Federation. There’s no other viable solution for composable architectures at scale.”
Sr. Director of Engineering at Varo
The supergraph has also helped Varo rapidly deliver on new initiatives since the launch of Varo Bank. Kenny McGarvey, GraphQL Lead at Varo, estimates that engineering velocity is 4x what it would be without the supergraph.
“There are a lot of different ways that the supergraph architecture makes us more efficient,” says McGarvey. “So many bugs that might have been deployed have been avoided because of Apollo’s managed federation model. You don’t have to worry about the order in which you deploy subgraphs, and that’s very important when you have 88 plus microservices on the back end. There is no other instance where you can say, yeah, go ahead and deploy our 12 services in any order you like. We don’t care.”
“I save at least 10-15 minutes per deployment — and that’s not even including our pre-production environments,” McGarvey estimates. With an average of 30 deployments per week and considering the average hourly cost of a software engineer in San Francisco, that equates to roughly $42K savings per year for just one deployment scenario and more than $132K in savings and off-sets considering preproduction environments and decoupled teams.
Technology and rapid innovation have always been a core component of Varo’s strategy, but as a bank, compliance with tight industry regulations is also a constant top priority for Varma and his teams. “Our customers need to know that their money is available to them whenever they need it, and they need to know that we will keep it safe for them until that time. So security is always a top priority for us, and additionally, there are of course constantly changing regulations that we must comply with,” Varma explains.
The supergraph and Apollo Studio give Varo an unparalleled ability to adapt and comply with these regulations without slowing down their product innovation. Sundar recalls, “The legal team would come to us and say that we need to implement a change that actually impacts our application logic. For example, we needed to change the selected auto-payment day from a 15-day window.
“Using a different API architecture, it could take quite some time to make that type of a change. With the graph, we are able to abstract that logic from our client applications and into the graph so that we can quickly make changes to it. And because all of our client teams use the graph, those changes will automatically be reflected across every client application without disrupting engineering workflows or introducing breaking changes.”
“We want to take every possible measure to protect our customers, and the supergraph helps us do that in a safe and scalable way.”
GraphQL Lead at Varo
The graph also helps Varo with fraud mitigation in addition to streamlining compliance. “We want to take every possible measure to protect our customers, and the supergraph helps us do that in a safe and scalable way,” McGarvey explains. “If there are certain risky transactions that we want to block or show an error message for since there could be a chance of fraud, we can put that logic right into the supergraph layer and control it across all of our client applications at once.”
The graph is at the center of Varo’s technology-driven approach to disrupting the banking industry. Varma explains – “technology is a big part of Varo, but ultimately our mission is to create a consumer-centric bank. And what consumers need from a bank is trust – trust that we will keep their money secure and trust that their money will be available to them at any time when they need it.
“The graph plays an integral part in how we build that trust with our customers. It’s the orchestration layer that allows us to build a best-in-class consumer experience, and it allows us to deliver that experience to our customers on their devices 24-7 so that they always have access to their money.”
If you’d like to learn about Varo Bank and its offerings, visit varomoney.com.
Varo is an all-digital bank designed to help consumers save money and improve their financial health.
When launching their new bank, Varo needed to choose a stack that would flexible enough for rapid innovation, yet secure enough for a bank.
By building on the supergraph, Varo was able to decouple backend teams and rapidly deliver a customer-centric banking solution that differentiated Varo in a crowded market.
How We Built a Bank on GraphQL
At Varo, our mission to build “a bank for all of us” began with building a modern...
Panel: Driving Transformative Change with the Graph
Engineering leaders increasingly see learning GraphQL as a strategic advantage to overcome the...