Data Science or Software Engineering: Which Should You Learn?

McTaba runs coding and data bootcamps in Kenya. This guide is written to be useful whether or not you ever join one.

Picture two people at neighbouring desks in Westlands. Both have a laptop, a code editor and a cup of tea going cold. One is fixing why a customer's M-Pesa payment did not mark an order as paid. The other is working out why the same customers stop buying after their second order.

The first is a software engineer. The second is a data scientist. Both write Python or something close to it, and both read other people's code. The difference is what they hand over at the end of the week.

A software engineer hands over a working thing: a checkout, an app, a USSD menu. A data scientist hands over an answer: a number, a chart or a model that predicts something. Most of this page is about which of those two hand-overs you would enjoy making.

We run one programme for each, so we know both sides of the desk. The bigger map of data careers is in the data science career map for Kenya. This page only settles the first fork.

Take the same imaginary Nairobi lender, a mobile loan app with a few thousand borrowers. Here is a Tuesday for each role.

Hour	Software engineer	Data scientist
Morning	Reads a bug report: a borrower repaid through the paybill and the app still shows the loan as due	Opens last month's repayments and checks how many are missing a phone number or have two date formats
Midday	Writes the fix, adds a test, asks a colleague to review the change	Compares borrowers who repaid late with those who repaid on time, and charts what differs
Afternoon	Ships the fix to the live app and watches the error logs	Trains a model on past loans and measures how often it is wrong on loans it has never seen
Last hour	Updates the ticket so support can tell the borrower it is sorted	Writes three sentences for the credit manager: what the model says and where not to trust it

Notice who waits on whom. The engineer waits on a clear spec and a working system. The data scientist waits on clean data, and spends more of the week than most beginners expect on getting it.

The two jobs also meet. A model nobody can call is a notebook, so someone has to put it behind an API. That is software engineering again. If you want to see where the lines blur, read how to deploy a model with FastAPI.

For the engineering side of the same day, what a software developer does all day in Kenya goes hour by hour.

If the idea of owning a live product, with real users and real outages, excites you, our Data Science to AI Engineering programme is probably not your first stop. The Marathon is. It sits at our coding bootcamp in Nairobi.

Software engineering leans on logic more than on maths. Loops, conditions, and breaking a messy request into small steps. School algebra helps. Calculus rarely appears in a payments app.

Data science asks for more. You meet averages and spread in week one, then probability, sampling and the idea of a fitted line. Later come the models, which are maths wrapped in code.

That does not mean a degree. Our Data Science to AI Engineering programme puts maths in its own phase, "Math, Statistics and Time Series", in weeks 7 to 8, and teaches it through code. No coding, data or maths background needed. If this is your worry, read whether you can learn data science without a maths background.

For software engineering, the same question has its own answer in whether you need maths to code.

Count the adverts, because adverts are where hiring shows up. BrighterMonday showed 241 jobs for "software developer", 41 for "data analyst" and 17 for "data scientist" (software developer search, data analyst search, data scientist search, checked on 3 November 2026).

Searches are blunt tools. "Software developer" also catches interns and hybrid roles, and one employer may post the same role twice. Read the numbers as a rough ratio, not a census. Even so, the gap is wide enough to matter: there are far more places to apply as a software developer.

The other pattern is how few adverts show pay. All 17 data scientist adverts showed "Confidential" where the salary goes (BrighterMonday data scientist search, checked on 3 November 2026). For figures from surveys, with their sample sizes, see what a software developer earns in Kenya and what a data scientist earns in Kenya.

Entry routes

Software engineering takes beginners more readily. A small portfolio of working apps, a GitHub account and one deployed project can open a junior interview. See how software engineer, web developer and IT roles differ before you aim at a title.

Data science often starts one step sideways. Many people enter as data analysts, using Excel, SQL and Power BI, then add Python and machine learning. Employers in banks and NGOs also ask more often for a degree. How a data analyst, a data scientist and a data engineer differ in Kenya maps the steps.

Neither field has a licence or a gatekeeper. Employers judge what you can show. No school can promise you a job in either, and any school that does should lose your trust.

We teach both, so here are the facts of each in one place.

	The Marathon (Fullstack Software Engineering)	Data Science to AI Engineering
Price	KES 120,000, or KES 100,000 if paid in full upfront	KES 140,000
Length	4 months full time or 6 months part time	5 months full time or 8 months part time
Next intake	11 January 2027	11 January 2027
Part-time hours	Monday to Thursday, 7 PM to 9 PM	four sessions a week, 7 PM to 9 PM
What you ship	Web apps and payments, built on the African Stack	m-pesa fraud detection, credit scoring, customer segmentation, rag support assistant (capstone options)

The Data Science programme runs 9 phases and 24 modules. Its first cohort starts on 11 January 2027, so we have no graduate outcomes to show for it yet. Hybrid: live online sessions on weekdays and in person sessions in Nairobi on Saturdays, all recorded.

The Marathon is hybrid when taken full time and fully remote when taken part time. Cohorts are small, at 10 learners.

The two overlap by design. The Data Science programme starts with Python and SQL, and it ends with models behind FastAPI. A graduate of either can read the other's code.

Answer without overthinking. Count how many lean each way.

1. You get a Safaricom statement with 400 rows. Do you want to find the pattern in it (data science), or build an app that reads any statement a user uploads (software engineering)?
2. When something works, which feels better? A page a stranger can use (software engineering), or a chart that makes a manager change a decision (data science)?
3. Uncertainty. Can you live with "the model is right about 8 times in 10"? Data science is full of that. Software engineering wants the checkout to work every time.
4. Cleaning. Does fixing 40 badly typed phone numbers sound like a puzzle or a punishment? Data scientists do a lot of it.
5. The job hunt. Do you want the widest field of adverts now (software engineering), or are you happy to enter as an analyst and climb (data science)?

Four or five answers on one side, and your choice is made. A split result means either path suits you. Start with the one you can begin this month, and add the other later.

Still unsure? Learn Python first. It is the first skill on both paths, and the Python a data science beginner needs costs nothing but time. A few weeks in, you will know which half you reach for.

If your answers lean towards questions and patterns, read the data science roadmap for beginners, then look at the Data Science to AI Engineering programme. It takes 5 months full time or 8 months part time and costs KES 140,000.

If they lean towards building, look at the Marathon in Nairobi: KES 120,000, or KES 100,000 if paid in full upfront, 4 months full time or 6 months part time, next intake 11 January 2027.

If you want to compare schools before you decide, read which data science course in Kenya suits you. Nothing above is a promise of a job or a salary. It is a way to pick the daily work you would put up with on a bad Monday.