Three Years at Cellulant - My 2025 Career Recap
1. Combobulation at the Beginning
If 2025 were to be a movie, the classic, Shoot ‘Em Up, captures my career experience. It started as planned, adhering to the roadmap in Q1.
Q2 rolled in smoothly, and then things went from 0 to 100 very fast. Interrupted by high-stakes, high-priority projects. Urgent regulatory requirements, tier 1 customers and everything in-between.
In order to meet deadlines, we embarked on hackathons upon hackathons and out goes my 2025’s New Year’s resolution.
Evidence is the fact that my last publication was in May of this year. I even have a draft I started in June that’s yet to be published. It was that busy!
It was not all for nothing. We recorded major wins, moved the limits of possibility. I am grateful to God for the opportunity to contribute and make an impact.
Follow me in this article as I recount some of the wins and lessons of my career in 2025.
2. What Went Well
2.1 Tactical Tech Debt Reduction
One of the major wins of the year was the improvement of our settlement system for scalability.
Every system is scalable until it gets punched by big data.
I am talking about hundreds of millions of records.
One example of the impact is reducing processing time for one of the biggest Telcos in East Africa by 86%.
The difference was like night and day. No more babysitting systems and praying to the server gods.
What we did differently is to go back to the first principles - how should this thing work?
We ended up with a highly scalable system that breaks down a single process into tiny bits that can be processed in parallel, and the results are combined for further processing.
My tip for you, dear friend. For consumer applications, apps that get messages from RabbitMQ, optimise for short processing time.
The time taken to process a single message should be in seconds, as opposed to long-running processes. If you have a lot of messages and each takes seconds to process, you can scale the number of consumers accordingly.
However, if one message takes hours, even if the queue backlog is in hundreds, things like app restart, timeouts, will result in longer overall processing time.
We did not stop there. We squeezed in some modernisation of our APIs that resulted in the capability to process transactions 3.6x faster, reduced operating cost, and improved our uptime 💪.
Another area where we made a major improvement is our back-office automations and processes. In case you don’t know, back-office operations can make or break a FinTech organization.
In our case, we built automations that improved onboarding, financial operations, audit, and regulatory compliance.
These set Cellulant on the path of becoming the best payment company in Africa.
Tech debt is the real definition of a dilemma. On one hand, we need to take care of it so we can deliver better service, and on the other, we need to build new features for new customers and revenue.
Next year, we take a stab at another section and continue until we win, and win absolutely.
2.2 Open API Spec
We started with the adoption of the Open API spec for our APIs. Integrating it into the Spring Boot ecosystem was straightforward and pleasant.
Why did we adopt it? This is because the Open API spec gives us API request validation, auto-generated controller classes, request and response models out of the box.
It also eliminates the need to manually update API documentation.
This is because every time we change the spec file, the models, generated classes, and the documentation are updated automatically.
We now plugged this into a RapiDoc template, giving us a beautiful UI that’s always up-to-date on every refresh.
You should try it too. I highly recommend it.
2.3 Wedding, Babies, and Promotions
As usual, to God’s glory, we had a team member’s wedding this year. We also had four newborn babies across Africa and India 👶.
It is always something I look forward to every year. To celebrate the welcoming of new babies and team members getting married. Want one? Consider joining my team 😁.
My team grew with new members joining us from Nigeria, Kenya, and India. I always count it a blessing to be working with distributed teams across different countries in Africa and India.
I also got a product manager, finally. It was not easy being the Engineering Manager and the Product Manager for such a large team.
So I am excited to get the much-needed help in this regard.
This year, 6 team members got promoted, including one person moving to become an Engineering manager and having their own team.
I considered this one of the best moments of my career.
You remember that my manager, Mike? He’s now the boss of my boss 🥳. A big congratulations to him, it is well deserved and a long time coming.
3. Lessons Learnt
3.1 Customer Feedback is King
One major lesson I learnt this year is to engage and involve your customer as much as possible while building.
Delivering work that ends up not being used is real. It wastes company resources, demoralises the team, it is bad in every dimension.
In one instance, we had to go back to the drawing board following feedback from a demo. I remember saying to myself, “Finally, we just demo this feature and then get the much-needed rest”.
Only for the customer to reject it on the demo day because it will create more problems than it will solve 😥.
Back to the drawing board we went, and out the window went the imagination of “resting after this”.
Anyways, please make sure you get the requirements well and let your customer know the “innovation” you’re planning before you implement it.
Na “innovation” sent us back to the drawing board.
3.2 Information Distortion
I observed another phenomenon that can lead to wasted Engineering work. It’s the information distortion.
You see, between what the customer says to the account manager, flowing to product, Excos, and finally getting to Engineering.
There are a lot of interpretations in between with, some creative liberties.
This is why the customer will ask for a Cat, and the team will deliver a Catalytic Converter.
Because the original problem statement has morphed in the process of interpretation.
So, as much as possible, do in-depth product discovery and speak with the customer during the build process.
3.3 Time Management and Planning
Have you ever heard it before that someone owed the company leave days? Now you have.
For the most part of the year, my leave days were in the negative. So, the lesson learnt here is to manage your leave days properly.
I guess I spent all my leave days in January and did not pay close attention to it. Thank God for public holidays 😀. In 2026, I am taking charge of my leave days.
In the same vein, 2025 was filled with high-priority, super-urgent projects that caused us to hunker down and go into hackathon mode to deliver them on time.
In retrospect, we could have planned better, reduced the number of things we needed to do and identified the high priority of high priority.
The lesson here is to start planning early and stick to the plan.
I poured so much into projects and meeting deadlines, I did not have enough time for things like writing, creating videos or even maintaining my open-source libraries.
The lesson for me this year is to be brutal about planning and not underestimate projects. Though we adopt the route of MVP, I am looking forward to an even leaner MVP.
In fact, the first question is, can this be solved without coding? 😆.
3.4 Responsibility, Ownership, and Growth Mindset
I observed two attributes that can make an Engineer stand out.
The first one is an insane sense of responsibility. This sense of “what has been placed in my care must go well”, is a rare trait.
This sense of ownership will lead to proactiveness, and proactiveness will lead to consistency, and consistency will produce reliability.
A reliable person will be given even more responsibilities and attending reward.
I remember earlier in the year while, we were interviewing candidates for the Senior Software Engineer role.
We did not proceed with a candidate because we observed they did not take any initiative of their own in their current place.
If you remember the story of a master that gave his servants talents while he was travelling. The guy that delivered the least simply lack sense of responsibility and ownership.
Because the poor servant did not see what has been handed over to him as his, he did the barest minimum by burying what was entrusted to him.
Do you know that one guy in the company that everyone misses once they’re on leave? That’s most likely someone with a high sense of ownership.
The lesson here is to be that person in your organisation.
The second trait is the growth mindset. You show some Engineers how to do something, you still need to repeat it over and over again. But you see, for some rare gems, you show them once.
They immediately adapt, adopt and make it their own, to the point of improving it.
That spirit of continuous learning. Satisfying the standards to the point that you are improving on them. It’s a desirable trait, and that’s how people get promoted among peers.
Think about it, someone was asked to complete a task, following certain standards, but they did not do it well. Instead, they’re acting as if “I am not allowed to be creative in the team”.
What type of creativity are we talking about here? Switch our programming language overnight? Vibe coding?
What does creativity actually mean for an Engineering team? How do standards and best practices move the organisation forward? That’s a topic for another day.
4. 2026 and Beyond
One of my goals in 2026 is to create space and observe holidays properly (reads not to work on holidays) 🤞.
I am also working towards expanding my team’s “social time” to include newer activities.
I also plan to complete my ongoing leadership course, and work towards writing my second book or something of similar magnitude.
I would like to appreciate all of my team members who were there through the thick and thin of 2025. Thank you so much for your contributions, and cheers to a productive 2026.
Merry Christmas and Happy New Year 2026 🎉
Cheers to making impacts and being a blessing 🥂
