Revolutionising and taking the world of payment processing by storm with MongoDB Atlas

When you make a payment in a store or coffee shop, tapping your card, waiting for it to go through, and taking your receipt is the standard experience. As shoppers, we barely notice it. But for small businesses, putting the shopper experience first comes at a price. Think slow payment authorizations, confusing bills, lengthy onboarding, and a long wait until funds finally appear in their bank accounts.

Dojo is the UK’s newest payment processing business. After just over two years, it has 100,000 users, amounting to a 15% share of the SME market. Why? Because its founders spent 15 years witnessing SME pain points at parent company, Paymentsense, and they designed a brand-new solution that’s twice as fast, takes minutes to set up, and swaps Z reports — the summary of the day’s transactions — typically printed on small rolls of paper, for an easy-to-use app.

“Paymentsense is one of the leading players in the industry, but we competed on price, not product,” says Nick Fryer, CTO at Dojo. “We wanted to solve the problems in the market and knew we could do something special at Dojo, building with our customers in mind and helping them thrive.”

Processing payments is a complex business. The entire system is event driven and relies on fast, highly secure interactions between multiple third parties. It’s also a heavily regulated industry, and one traditionally mired with legacy technology.

“New payment solutions are not usually built on new infrastructure. We wanted to break the mould and do something totally different,” Fryer reveals. “Our environment is built for agility. It’s multi-cloud so we can take advantage of the best cloud services for different use cases. Rather than hosting legacy software on the cloud, we build using microservices so we can release new features quickly and often.”

When Dojo was looking for a database to support microservices, it wanted a managed service that was scalable, fast, reliable, and compatible with multiple clouds across a variety of geographical regions. It also needed a solution that was compatible with its preferred programming languages, C# and Go (also known as Golang).

The team selected MongoDB Atlas and moved around 1,000 microservices onto the platform. This supports processes around sales, services, and onboarding, and the platform is also used for reconciliation.

“Any processes that need fast serialization, deserialization of JSON objects, or involving multiple JSON objects work best on NoSQL databases,” explains Fryer. “MongoDB is super fast and super flexible. Settlement, clearing, and billing run on AWS, Oracle Cloud Kubernetes Engine, and Google Cloud. They all use different databases, but publish to a managed Kafka bus that is consolidated in MongoDB Atlas to make sure everything adds up.”

Nick Fryer, CTO at Dojo