Build journal
Another project?
I already had enough on my plate. I built KUMBA anyway.
Ndiaga Ndiaye4 min read
Build journalShare on LinkedIn
Build journal
I already had enough on my plate. I built KUMBA anyway.
It was supposed to be a news website.
Nothing too complicated: collect stories, organise them, publish them. Use some automation and AI so I wouldn't have to spend my days copying and pasting articles.
Well, “nothing too complicated” is something I tend to say before I start a project. 😅
I already have a day job, client work, and KAMIT, the small company I'm trying to grow. My calendar wasn't exactly asking for another adventure.
And yet I launched KUMBA MEDIA.
Why? Good question.
The original idea was straightforward: one place to catch up on African news.
But while building it, I started thinking about what was missing whenever I looked something up.
Imagine a company announces a major investment in Senegal. I can read the story, close the tab and move on.
Or I might have more questions:
The answers often exist. They're just scattered across different sites.
So I began imagining a platform that wouldn't simply stack articles on top of one another. It would connect the information.
That's how a “small news website” started turning into something much more ambitious.
Was that necessary? Probably not. Was the developer part of my brain excited? Absolutely.
Automation is at the heart of KUMBA.
I want the platform to collect, process, classify and publish information without requiring me to approve every single step manually.
Not because AI is magic.
It's because one person can't monitor hundreds of sources, clean the data, connect the dots and publish around the clock.
But there's a catch: automating a bad process is a fantastic way to make mistakes faster.
You might publish the same story ten times. Get a date wrong. Confuse an official announcement with a rumour. Or generate something technically coherent that nobody would actually want to read.
So the hard part isn't simply getting automation to work.
It's making it useful and trustworthy: careful source selection, duplicate detection, automated checks, transparency and constant refinement.
I'm still figuring out how to get all of that right.
Here's something I keep noticing about myself: when I have an idea for a feature, my first instinct is to work out how to build it.
The database? I have a plan.
The interface? I can already picture it.
The API? We'll figure that out.
But there's one question that is much less fun:
Does anyone actually need this?
That question can save weeks of work. And developers are sometimes very good at postponing it.
KUMBA keeps bringing me back to it.
News is the first building block. I'm also exploring opportunities listings and, eventually, a company directory.
But another feature isn't valuable just because it exists. It needs to help someone find something, understand something, or make a better decision.
Otherwise I've simply built a longer navigation menu. A lovely one, no doubt. But longer. 😂
I tend to work across several areas: software development, design, digital communications, SEO and project management.
With KUMBA, they all meet in one place.
I need to think about data architecture, how information is presented, how people find it, what can go wrong and what visitors actually gain from using it.
That's exciting.
It's also an excellent way to make life more complicated than necessary.
To keep myself from running in every direction, I try to follow one simple loop:
Build → ship → observe → fix.
Then repeat.
Not every new feature needs to be called a revolution. Some will survive, some will change and others will probably end up in the bin.
That's part of building a product.
I could end this post by promising that KUMBA will transform access to information across Africa.
That would sound impressive.
Unfortunately, I don't know if it will.
KUMBA is still young. I have a direction, features I'm working on and plenty of unanswered questions.
I want to go beyond a news feed and eventually offer a place to discover companies, opportunities and information linked together.
I don't yet know exactly what shape that will take or how long it will take to get there.
So, for now, I keep building.
And if things don't go according to plan, at least I'll have something interesting to write about here. 😉
This post is part of my build journal: unfinished projects, changing ideas and lessons learned along the way.
