What Kampala Tech Interviews Really Test (Not Your Code)

A backend developer with three years at a logistics startup once told me he solved the take-home assignment in ninety minutes, submitted it a full day early, and heard nothing for two weeks. The candidate who got the job submitted a messier solution, twenty minutes before the deadline, with one function that didn’t even run. She got an offer. He got a form rejection. The difference wasn’t skill. It was what each of them put in the README.

That’s the part almost nobody preps for in Uganda’s tech hiring, and it’s costing capable developers, network engineers and IT support staff offers they’re qualified for.

The take-home test isn’t testing your code

Most mid-sized tech employers in Kampala, from fintech startups in Ntinda to the engineering teams inside telecoms like MTN Uganda, now screen with a take-home assignment before anyone sits across a table. Candidates treat it as a coding exam. It’s actually a communication exam wearing a coding costume.

The woman who got the offer above wrote four sentences at the top of her README explaining what she’d skipped, why, and what she’d have built with another day. The reviewer didn’t need perfect code. He needed proof she’d think clearly about trade-offs on a live production system where “I ran out of time” happens every single sprint. The guy who submitted clean, complete code with zero commentary gave the panel nothing to judge except the output, and output alone rarely wins a close call.

If you take one thing from this article, take this: document your decisions, not just your deliverables. A rough solution with sharp reasoning beats a polished one with none.

Three interview formats, and candidates keep mixing them up

Tech hiring in Uganda isn’t one format. It’s at least three, and preparing for the wrong one is a common, avoidable way to lose a role you were capable of doing.

• The whiteboard/live-coding round. Common at larger employers and outsourcing shops serving international clients. Panels want to see your process out loud, not a silent genius. Narrate your thinking even when you’re stuck. A candidate who says “I’d check if this list is sorted before picking an algorithm” out loud scores higher than one who stares at the screen and produces a correct answer thirty seconds later in silence.
• The systems/scenario interview. Used for network engineers, IT support leads and infrastructure roles, often phrased as “walk me through how you’d handle a branch-wide outage at 8am on a Monday.” This is a judgment test, not a trivia test. They want your triage order, not a list of every protocol you know.
• The take-home plus debrief. Described above. The debrief conversation afterward matters as much as the code itself, because it’s where you explain the parts you’d redo.

Ask upfront, when you’re invited, which format you’re walking into. It’s a completely normal question, and it tells the recruiter you’ve done this before.

There’s a fourth format worth naming separately, because it’s spreading fast through Kampala’s smaller product teams and outsourcing houses: the pair-programming session, where you and an engineer on the team build something together over a video call for forty-five minutes. Candidates who’ve only ever coded alone freeze here, not because the problem is hard, but because they’ve never had to think out loud while someone watches their cursor move. Practise this specifically. Grab a friend, share a screen, and solve a small problem together before your first real one. The skill being tested isn’t your syntax. It’s whether you’d be bearable to sit next to during a real sprint.

What your GitHub says before you say anything

Recruiters at Kampala’s better-paying tech employers now check a candidate’s GitHub profile before the first call, and most candidates’ profiles actively work against them. A profile with eleven forked tutorial repos and nothing original tells a reviewer you’ve followed instructions well and built nothing of your own. One real project, even a small one, with a clear README, a handful of commits spread across actual working sessions rather than one dump at 2am the night before an application, and a live demo link if the project supports one, outperforms a CV bullet point every time. If you’re job hunting and your public profile is empty or stale, fixing that costs you a weekend and changes how the first five minutes of your interview go, because the interviewer walks in already curious rather than neutral.

The same logic applies to candidates applying for infrastructure or support roles who don’t write much code day to day. A short, documented home-lab project, even something as modest as setting up a monitoring dashboard for a Raspberry Pi, signals curiosity that a CV line never will.

The skills gap nobody mentions in the job ad

Uganda’s tech sector has a specific, well-documented shortage: strong junior coders, weak production-readiness. Bootcamp and university graduates can write functioning code but have rarely touched version control discipline, basic testing, or how to read a stack trace from a system they didn’t write. Employers know this, so panels now probe for it directly, usually with a single blunt question: “Tell me about a bug you introduced that broke something, and how you found it.”

There’s no right bug to describe. There is a wrong answer, which is claiming you’ve never broken anything. Panels read that as either inexperience or dishonesty, and neither helps you. A specific story, even an embarrassing one, where you traced a null pointer or a race condition back to your own commit, signals you can be trusted near a live system. That’s worth more to a hiring manager than a perfect CV line about “strong problem-solving skills.”

Certifications that actually move the needle, and ones that don’t

Candidates spend money on certifications that don’t change a shortlist decision and skip the ones that do. For cloud and infrastructure roles, an AWS certification carries real weight with employers running anything on cloud infrastructure, partly because it’s verifiable in thirty seconds and partly because it proves you’ve touched real tooling, not just theory. CompTIA Network+ still opens doors for IT support and helpdesk roles at banks and telecoms, where the job is diagnosing a faulty switch at a branch in Mbarara, not architecting a cluster.

What doesn’t move the needle as much as candidates hope: generic “introduction to programming” certificates from platforms nobody in the hiring panel has heard of. They’re not worthless for learning. They’re close to invisible on a CV next to a GitHub profile with actual commits. If you’re choosing where to spend limited money or time, a public portfolio of real, working projects beats a stack of unverifiable certificates almost every time.

What panels at telecoms are actually hunting for

Engineering and IT roles at Airtel Uganda and similar large telecom employers run differently from a ten-person startup. The teams are bigger, the systems are older and more interdependent, and a mistake has a bigger blast radius: a misconfigured update can affect customers across a whole region, not just one app’s users. Panels there lean harder on process discipline. Expect questions about how you’d roll out a change safely, who you’d notify before touching a shared system, and how you’d roll back if something broke at 2am. If your only experience is solo projects where you were the only person who’d ever see the code, say so honestly and explain how you’d adapt, rather than pretending you’ve run change-management processes you haven’t.

This is also where Uganda’s regulatory backdrop occasionally comes up in interview, more often than candidates expect. Telecom infrastructure roles sit under oversight from the Uganda Communications Commission, and government digital-service roles increasingly route through standards set by the National Information Technology Authority. You don’t need to be an expert in either body’s regulations. Knowing they exist, and that uptime and data-handling aren’t just internal company preferences but things regulators care about, is enough to not look blindsided if the topic surfaces.

Salary expectations, stated plainly

Junior developer and IT support roles in Kampala typically start somewhere in the UGX 1.2 to 2 million range, with mid-level backend, network or systems engineers often landing between UGX 2.5 and 5 million depending on the employer, the stack, and whether the company serves international clients who pay closer to global rates. Senior engineers and specialists in cloud, security or data roles can clear that range substantially, particularly at telecoms, banks’ technology divisions, or outsourcing firms billing in dollars. These are planning ranges, not guarantees, and they move with the specific employer’s size and client base. Negotiate from a number you can defend, not a number you hope for.

Before you walk in

Build a short, honest story about one real failure you fixed, mapped to the kind of system the role touches. Pin down which of the interview formats above you’re facing and prepare for that one specifically, not a generic mix of all of them. If the role is a career pivot into tech from a different field, our piece on switching into cybersecurity from a non-technical background covers how to frame that transition credibly. And if you make it to an offer, don’t take the first number without reading our guide to negotiating a salary offer in Uganda first.

For the wider picture of who’s actually hiring across Uganda’s ICT sector right now, our industry guide to tech and ICT careers in Uganda is the right next stop. When you’re ready to apply, current openings across Uganda’s tech, telecom and IT employers are listed on Kampala Index’s jobs board.