Learn Data Science While Working: A Plan for Nairobi Professionals

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

Picture a Tuesday at 6:40 PM on Thika Road. The matatu is stuck at the Roysambu flyover, your laptop is in your bag, and you are meant to start a class at 7. Anyone who has tried to study after work knows this scene.

Learning data science with a full-time job is possible. People do it. What decides the outcome is rarely brains. It is whether you protect a fixed block of hours each week and whether the plan survives a bad month.

This guide gives you three time budgets, the facts of our part-time track, a way to ask your employer to pay, and what to practise on your own company's data without breaking the Data Protection Act. The wider map of the field is in the data science career map for Kenya.

Pick the budget you can keep for six months, not the one that sounds brave. Here are three, with what each realistically gets you.

Budget	Pattern	What it fits
5 hours a week	Three evenings of 1 hour, plus a 2-hour Saturday morning	Self-study at a slow pace: Excel, SQL and basic Python. Progress is steady but slow.
8 to 10 hours a week	Four weekday evenings of 2 hours, plus a short weekend review	A structured part-time track. This is the pattern our own part-time schedule assumes.
15 hours a week	Evenings plus one full Saturday	A faster self-study pace, or a part-time track plus extra project time. Hard on family and rest.

Two rules apply to all three. First, put the hours in your calendar like a meeting, with a start time. "When I have time" is not a plan. Second, count from the day you start the budget, not from your best week. A budget that only works in a quiet month fails in the first busy one.

The weekly rhythm matters more than the length of any single session. A 30-minute practice on a weeknight beats a 5-hour Sunday that you skip every second week. Short, regular sessions also fit the way code is learned: you type, you break it, you fix it, you remember it.

Where to spend the hours depends on your stage. Beginners should follow the data science roadmap for beginners, which sets the order. If you already live in Excel at work, how to move from Excel to Python is the fastest first step.

If you are still deciding whether data is your direction at all, read whether to learn data science or software engineering first, and save yourself six months.

We run our Data Science to AI Engineering programme in two shapes. Full time takes 5 months. Part time takes 8 months, with four sessions a week, 7 PM to 9 PM.

That is 4 sessions of 2 hours, or 8 hours in class every week. Add practice and the weekly project, and plan on a few more hours. The programme has 24 weekly projects, so something is due most weeks.

Hybrid: live online sessions on weekdays and in person sessions in Nairobi on Saturdays, all recorded. Because all sessions are recorded, a missed evening is not lost, though you still have to catch up before the next one builds on it.

The next cohort starts on 11 January 2027. The programme costs KES 140,000. Imagine spreading it evenly over the 8 months of the part-time track: about KES 17,500 a month, for example, before any instalment plan. Instalments available, confirmed at onboarding.

No coding, data or maths background needed. A laptop and internet. Labs run on Google Colab and Kaggle.

See the phase list and the cohort details on the Data Science to AI Engineering programme page. We have no graduate outcomes to report, because the first cohort starts in 11 January 2027.

Honest warning: part time is not the easy option. Eight months of evenings tests your energy more than a four-month sprint tests your savings. Decide which you can sustain.

Many employers pay for training that helps the business. A request that names a business problem has a better chance than "I want to learn data science".

Build the request in four short parts.

1. The problem. Name one task your team does by hand that data skills would speed up. For example, a monthly stock report that takes two people two days.
2. The plan. Say which skills you will learn, in what order, and what you will deliver first. A first deliverable by month three keeps it concrete.
3. The cost and the hours. State the fee and the evening schedule. Make clear that class hours fall outside work hours, if that is true, and ask for one thing only: payment, or a flexible leave day for the final project.
4. The commitment. Offer to share what you learn with the team, and, if the company wants it, to stay for an agreed period.

Some employers prefer to split the fee. Others reimburse after completion. Ask which applies before you enrol, and get the answer in writing. Ask us for an invoice in the company's name if the finance team needs one.

If they say no, you lose nothing by asking. You have also told your manager you want to grow, which sometimes leads to a different yes later. For a small team that cannot spare you, a different offer may work better: corporate training in data analytics teaches a whole department at once.

Do not wait for sponsorship to start. Use the free steps first and bring a finished first project to the conversation.

Your office holds the best practice data you will ever find. Real sales, real mistakes, real mess. It is also the data most likely to land you in trouble if you handle it badly.

Good first projects on work data:

• A report you already produce. Rebuild it in SQL or Python, so it refreshes in minutes instead of hours.
• A data quality check. Count blank phone numbers, duplicated customers and impossible dates in one table.
• A trend chart. Weekly sales or tickets, with the one odd week explained.
• A simple forecast.

Three cautions before you open a single file.

First, ask permission. Work data belongs to your employer. Tell your manager what you plan to do and why, and keep the answer in an email.

Second, mind personal data. Customer names, phone numbers, ID numbers and loan records are personal data under Kenya's Data Protection Act, 2019. Do not copy them to a personal laptop, a free notebook service or a GitHub repository. Remove or replace identifying fields first. How to anonymise personal data in Kenya explains how, and where the law draws the line.

Third, a portfolio must not leak. If an employer cannot be named, show the method on public or invented data, and say what you did at work in words.

Imagine you start in January with a job in operations, and you keep the 8-hour pattern. Here is a sensible shape.

• Months 1 and 2. Spreadsheets and SQL on your own spending, then on an anonymised work table.
• Months 3 and 4. Python and pandas. Clean one real file end to end.
• Month 5. Statistics and your first chart that changes someone's mind at work.
• Month 6. A small project you can show, and a CV line that names it.

Expect weeks that go wrong: a late meeting, a sick child, month-end at the office. Plan one "catch-up" week every six, with no new material. It keeps one bad week from turning into a lost month.

Keep a one-line log each night: what you tried, what broke. After three months it becomes your best interview material.

This week, block your hours and do one thing with your own data. If you want a cohort around you, look at the Data Science to AI Engineering programme: 5 months full time or 8 months part time, KES 140,000, next intake 11 January 2027.

If you are comparing schools first, read which data science course in Kenya suits you. No programme, ours included, can promise a job or a pay rise at the end. What a part-time route offers is a way to learn without giving up your salary while you do.