Nsano introduces new services 67% faster with MongoDB

THEIR CHALLENGE

Using MongoDB Atlas to enhance scalability and enable growth for Nsano

In the Twi dialect of Ghana’s Akan language, Nsano means, ‘fingertips.’ The word neatly articulates the eponymously named fintech’s mission: to make financial services readily available and accessible for all. Nsano was founded in 2013 and is headquartered in Ghana’s capital, Accra. Nsano specializes in digital payments processing, striving to democratize how e-money moves across Africa by reinventing financial service delivery with technology-driven solutions. Nsano's vision is to process 50% of Africa's GDP by 2030, offering frictionless, low-cost payment solutions that address multi-currency challenges, fraud, and data privacy. By partnering with banks, businesses, and financial institutions, Nsano aims to scale digital payments across borders.

The company’s range of services includes mobile money aggregations, cross-border remittance, and bulk transaction processing. They currently have footprints in over 30 African countries, and the UK, with regional offices in five African countries—Ghana, Liberia, Zambia, Cote d'Ivoire, and Sierra Leone—Nsano aims to cover the rest of Africa before 2025.

Christian Glouin, Chief Technology Officer at Nsano started as a software engineer and has witnessed its trajectory. “When we first started building the platform, we didn’t have a proper orientation as to where we were going to end, therefore having a fully structured database was out of the question,” he explained. “Content was rapidly changing, and we needed something that was flexible. That is what led us initially to MongoDB.”

The team started its journey with MongoDB Community 3.6. But as Nsano’s operations grew, they began to struggle with managing it themselves, and moved to a third-party managed instance of the database. Within a year there were issues. “The software changes and upgrades we were doing led to bottlenecks,” said Glouin. “For every 1,000 requests into the database, some were simply failing.” Other pain points were consistently slow reads and writes, considerable downtime, and failure to hit uptime SLA for customers. Moreover, with its existing platform Nsano was unable to scale, the team was self-managing security, and paying around £24,000 a year for a free version that could have been managed in-house.

OUR SOLUTION

Acquiring the flexibility to manage rapidly changing content

Frustrated by the challenges of using the third-party managed MongoDB Community 3.6 version, Nsano began to consider alternatives – at one stage looking to re-platform to a relational SQL-based solution. It was a fortuitously timed approach by MongoDB in December 2023 that resulted in them moving instead to MongoDB Atlas. “The reason we chose MongoDB from the get-go was because we needed the flexibility to add objects or details on the fly without having to change a system all the time,” said Glouin. Already familiar with MongoDB’s technology, if not specifically the MongoDB Atlas interface, Nsano were nevertheless pro-support customers, meaning they received four days of funded consultancy.

That arrangement helped accelerate the migration from Nsano’s legacy third-party managed MongoDB Community to a MongoDB Atlas supported version. “It was a straightforward operation,” said Glouin of the process, which began in May and concluded in July 2024. Any challenges that did exist were largely internal – code that needed to be optimized in some areas to ensure the database could be moved quickly, for example. After that, “things went very smoothly,” said Glouin. “I didn’t have to do a rollback, all the processing issues I was having with our legacy version just disappeared.” Occasionally Glouin receives alerts about queries that need to be optimized. “It’s insightful for us, because we can rely on the automated suggestions of the system to improve our services,” he added.

OUTCOME

Scaled database allows quick expansion of services in line with business vision

A major benefit for Nsano has been a significant reduction in the total cost of ownership (TCO). Previously, Nsano spent over £2,000 per month on their legacy solution; now, with MongoDB Atlas, costs are between £1,000 and £1,500—a reduction of around 25%.

Beyond cost savings, resource efficiencies have also been substantial. Nsano no longer requires a dedicated database administrator. "I rarely need anyone to manage the service," says Glouin. "It almost feels like it's auto-managing itself." Developers’ lives have also improved, as they no longer need to optimize or fix database issues. "They just write their code, add the connection details, and move on," Glouin adds.

Security has been another key benefit. Previously, Nsano’s data was not encrypted in transit, making it vulnerable. Now, with MongoDB Atlas’s strong security defaults, including encryption, Nsano is fully compliant with PCI DSS and other financial certifications, “without having to sweat that much,” says Glouin.

Expansion is now happening at a ‘terrific pace.’ Nsano’s previously monolithic application had three or four main services; after migrating, it now has fourteen. "We're using a microservices architecture," says Glouin. "We’re spinning new services every week." What used to take 6 weeks now takes just two weeks—a 67% acceleration. For example, the team quickly built a transaction switch that determines the route for financial operations. “MongoDB is brilliant for that,” says Glouin. “It eases development and deployment; within a week, we have enough data to decide whether to roll back or improve.”

Additionally, MongoDB Atlas's caching capabilities have allowed Nsano to reduce the cost of its caching layer and improve throughput responsiveness. “There are lots of things I had no idea MongoDB could do that we’ve started exploring,” says Glouin. “For me, it's like a treasure trove—like giving candy to kids!” To scale its services and realize its vision, Nsano first needed to 'sort out' its database. "Now that this has been done, building services is no longer a problem,” says Glouin. “Any new service requiring a database operation goes directly to MongoDB—it's like a walk in the park for us.”