What East Africa’s Banks Must Get Right

The same concern comes up in almost every conversation I have with a bank executive, whether in Nairobi, Kampala, Dar es Salaam, Kigali or Lusaka: we cannot afford to be left behind.

They are right to worry. Financial services in East Africa are competitive enough that technology increasingly determines who wins a customer and who quietly loses one. Expectations are higher, customers have more choice and tolerance for friction is disappearing.

The constraint is rarely budget or appetite. The harder problem is knowing what technology can realistically do for your business, and then knowing how to execute against that. Plenty of institutions have the first half. Far fewer have both. That gap is where transformation begins to stall.

AI is exposing it particularly clearly. A model is only as useful as the data it can reach and the systems it is allowed to act on. When customer records sit in different places that do not speak to each other, infrastructure cannot scale with demand, or every new requirement means another procurement cycle, the pilot never gets to become a product.

Getting past that point requires work that attracts far less attention than an AI announcement.

Data architecture, infrastructure, integration, governance, security, resilience and skills determine what an institution can actually build and sustain.

The infrastructure challenge has changed too. Banks are balancing regulation, latency, cost,

security, legacy investments and pressure to innovate faster. Public cloud may make sense for one workload, private infrastructure for another, while some systems remain on-premise. The question is no longer simply whether to move to the cloud. It is whether that mix can behave like one environment rather than five.

I have spent much of my four years at Computech working on this challenge as we built our cloud business from the ground up. The lesson has been consistent: complexity itself is not necessarily the problem. Poorly managed complexity is.

Resilience deserves the same attention. A bank makes a basic promise to its customer: the money is there and the service works. Every outage and every breach weakens that promise.

Security, recovery and business continuity are commercial questions because their consequences reach customers, revenue, regulators and reputation. Prevention matters, but so does what happens when prevention fails. How quickly can the bank recover, and how much of the business can continue while it does?

Passing The Test

AI ultimately has to face a similarly practical test. I would rather see an institution put a model behind fraud screening, customer service or credit decisioning and measure what it changed than announce an AI strategy whose commercial value cannot be explained.

Did fraud losses decline? Did decision times improve? Did the cost to serve fall? Did the

customer experience improve? These are the questions that determine whether an investment survives the next budget cycle.

Before joining Computech, I spent three years at Microsoft in the US. The technology available to a bank in Lusaka is much the same as the technology available to one in Seattle. The difference lies in what gets built with it.

Working with financial institutions across five African markets has reinforced that lesson. Technology cannot compensate indefinitely for fragmented processes, weak governance or unclear accountability. Great companies need great people and sound processes before technology can do what it is supposed to do.

For East Africa’s banks, getting ahead now requires more discipline than theatre. Know where the technology creates value. Build the foundations it needs. Put the right people and processes around it. Then execute.

My own rule is simple: under commit, overdeliver.

Banks have heard enough promises about transformation. What earns their trust is a system that works on the day you said it would.

This article was written by Hussein Popat, Head of Growth, Computech Limited