Follow Us on Google
By Katherine Diakpomere
A product manager at Access Bank, Adams Adeleke has argued that more data and more software will not retire the need to decide what a number actually means.
Nigerian banks have spent the past decade buying systems, hiring analysts and promising customers that the next application will make money move with less friction. The coronavirus months have only hurried that spending. What they have not retired is an older problem: someone still has to decide which process is worth automating, which customer record can be trusted, and what to do when the dashboard and the branch tell different stories.
Adams Gbolahan Adeleke works inside that problem. He is a product manager executive at Access Bank, a role he has held since March 2018. He trained as a petroleum engineer at the University of Ibadan, graduating in 2015, then spent July 2016 to January 2018 as a graduate business analyst at Shell Nigeria Exploration and Production Company. The oil campus taught him gap analysis. The bank is where he has had to apply it with a licence in the room.
The timing of the bank job is not incidental. In March 2019 a Federal High Court sanctioned Access Bank’s merger with Diamond Bank. The combined brand was launched on 1 April. Access Bank and contemporaneous reports, including Reuters, put the enlarged customer base at about 48.6 million, then among the largest in Africa by that measure. Adeleke’s documented work sat in the integration itself: extracting, transforming and loading customer data so two institutions could report, serve and remain recognisable to the supervisor as one. It is not the sort of assignment a bank gives to someone it is still testing. If the files do not match, the customer does not match.
He is careful, in conversation, not to confuse a working system with a solved problem. “Technology is most valuable when it solves a real problem. The ability to build something is important, but understanding the problem you are solving is what determines whether that solution will actually create value,” he said.
Other News
That distinction is easy to lose in a year when every vendor is selling artificial intelligence as a service. What Nigerian banks actually have, in 2020, is more modest and more useful: scoring models, fraud rules, process automation, and warehouses of customer data that grew again after the Diamond combination. Machine learning can flag a pattern. It does not tell a product owner whether the pattern is a new behaviour, a dirty feed or a queue at the ATM. “Data can tell you what is happening, but it does not always tell you what should happen next. That requires context, experience and judgement,” Adeleke said.
The World Economic Forum’s Future of Jobs Report 2018, still the Forum’s latest full edition this September, estimated that by 2022 no less than 54 per cent of employees would need significant reskilling. Analytical thinking and innovation were among the capabilities employers expected to rise, alongside what the report called human skills: critical thinking, creativity, complex problem-solving. Programming and technology design were on that list too. The Forum did not suggest that judgement would be the residue left after the software had finished. It suggested both would be demanded at once.
For young Nigerians the public advice remains to learn to code. It is not bad advice. Access Bank, like every large lender, needs people who can implement. It also needs people who can sit between a control function, a technology vendor and a customer journey and say which change is safe enough to ship. Adeleke’s path from Shell to a merger workstream is one local example of how that second skill is trained: not on a certificate alone, but on a live process that cannot be paused for a seminar.
There is a further, unfashionable implication for managers. Junior analysts still learn by preparing the tedious extract, checking the spreadsheet and sitting in the room while a more senior colleague decides. If those tasks are stripped out too quickly in the name of efficiency, the bank may discover it has automated the apprenticeship. Adeleke’s point is not that tools should be refused. It is that someone must remain accountable when the model is confident and the customer is not.
Nigeria makes that accountability heavier. Identity data is uneven. Connectivity fails. Customers share handsets. A scoring rule that looks clean in a slide deck can misread a market in which trust, cash and informal work still carry much of daily life. A technically neat product can fail because it misunderstands the user. A simpler one can succeed because its owners understood the constraint. Analytics supplies evidence. Software supplies speed. Judgement still has to connect both to the street.
He does not present himself as a prophet of the next platform. He presents a working method: understand the process, name the gap, stay with the result until operations can live with it. That method travelled from an energy graduate scheme into one of the country’s largest banking integrations. Whether it later travels into other sectors is a question for another year. For this one, inside Access Bank, it is already a job.
