💻

Data Scientist

Saintis Data · Teknologi IT

Starting
RM4,500 - RM6,500
Senior
RM12,000 - RM28,000
Entry
Degree
Short answerA data scientist builds models that predict things — which customers will leave, which transactions are fraud, next month's demand.
This is the most oversold career to Malaysian students, and this page is going to be honest about it. It pays well and it is interesting. It also needs more preparation than people say, and has far fewer entry-level posts than data analysis.
No licence.
What it actually requires: solid statistics (not just code), Python, and often a master's for a first post.

What does a Data Scientist do?

Builds predictive and machine learning models to solve business problems — statistics meets coding meets domain.

A day in the life

Data scientists build predictive models: customer churn, fraud, dynamic pricing, recommendation systems.
A typical day: model experiments (Python), feature engineering, measuring performance, presenting to business, working with engineers to productionise.
Half science, half communication.

Is this right for you?

A good fit if you: are strong in statistics & Python, enjoy repeated experiments (most fail), and can tell results as stories to management.
Less suitable if you: want certain answers — this field is full of uncertainty.

Salary & career ladder

Salary range
Junior data scientist: RM5,000–8,000.
Data scientist (3–5 years): RM9,000–16,000.
Senior: RM16,000–26,000.
Head of data science: RM25,000–45,000.
Banks, telcos, large e-commerce firms and technology MNCs pay at the top end.

Numbers to put side by side
The starting salary is higher than a data analyst's — but you may start two years later and with master's fees paid.
Do that calculation before choosing. Two years of an analyst's salary, plus two years of fees not spent, is a large sum to catch up on.

An advantage that does not exist in most other careers
Software work can be done from anywhere. A developer in Malaysia working remotely for a Singapore or American company is paid in SGD or USD while paying rent in ringgit.
The difference is large — large enough that it changes the whole arithmetic of this career, and it does not exist for a doctor, a lawyer or a civil engineer, whose licences are tied to a country.
But it is conditional on two things, and both can be built starting from SPM:
English at a real working level — not merely a pass, but enough to argue about system design in a meeting.
Work you can show — GitHub, projects that run, real contributions. Companies hiring remotely cannot interview you the usual way, so they lean on what they can see.
Students who build both while studying end up in an entirely different salary market from classmates who only collected certificates.

What AI changes — this career is in an odd position
Data scientists build AI. That does not make them immune to it.
Exposed: standard modelling, hyperparameter tuning, routine exploratory analysis, boilerplate. AutoML and AI tools now do much of this.
Not exposed: framing a business problem as a modellable one, noticing that the training data is biased, understanding when a model should not be trusted, and explaining a result to people who must act on it.
The real effect: "can call a library" is no longer a marketable skill. Statistics and judgement are, more than before.
That sharpens the advice on this page: the mathematical foundation matters more than it used to, not less.

City vs hometown

Concentrated in KL (banks, telcos, e-commerce, GLCs); Penang (smart manufacturing) is growing.
Regional remote is solid for the experienced.

Study path after SPM / UEC

Data Science / Maths / CS degreeThis page is going to tell the truth about this career
Data science has been marketed to Malaysian students as the hottest career in technology. Part of that is true: it pays well and the work is interesting.
The part usually left out:
Far fewer entry-level posts exist than in data analysis. Every company needs analysts; only companies large enough to have genuine prediction problems need data scientists.
A master's is often required — that is 1–2 extra years and extra fees before your first salary.
Statistics is the real barrier, not programming. Code can be learned; statistical judgement is harder.

The route we recommend, and why
SPM (Add Maths) → Statistics / Mathematics / Computer Science degree → data analyst first → master's or specialisation → data scientist.
Why go through analysis: you are paid while learning, you learn the actual business, and you find out whether you genuinely like data work before investing another two years in fees.
Students who jump straight in often find the posts they want require experience they cannot get.

What the job actually is
Less glamorous than expected. It is widely reported that a large share of the time goes on cleaning and preparing data rather than building models.
The modelling part is interesting. The part that decides whether you are useful is whether you notice when the data itself is broken.

Related fields
Data analyst — the recommended way in. Data engineer — more posts, often higher pay. Machine learning engineer — taking models to production, the engineering side. Economist — statistical modelling on different problems.If you are taking the UEC instead of SPM
Everything above still applies to you — the subjects are the same disciplines, only a different exam paper. What changes is the route after it.
The UEC is not accepted for direct entry into a Malaysian public university degree. That is why independent school students overwhelmingly go to private universities in Malaysia, or abroad.
The UEC is treated as equivalent to STPM and A-Level, and it is recognised in the UK, the United States, Canada, Australia and Taiwan — which is why the overseas rate from independent schools is so high.
And the part that costs families real money — read this one properly:
PTPTN eligibility runs through SPM, and a UEC alone does not carry it. To keep the loan available you need a complete SPM, which means two things people get wrong:
Sejarah must be passed. Since SPM 2013 a pass in Sejarah (minimum E) is compulsory for the certificate itself — fail it and you do not have a complete SPM at all.
Bahasa Melayu is usually required at credit (grade C), not merely a pass.
The institution and programme must also be PTPTN-recognised — check that on the PTPTN gateway before you commit to a college.
The UEC route pushes you towards a private degree, and PTPTN is what pays for it. If you have not sat SPM, sit it — and do not treat Sejarah as the throwaway paper.
These conditions change. Verify the current rules with PTPTN before relying on any of this.

Universities

UM, UKM, USM, UTM (data science), APU, MMU; local DS Master's programmes are multiplying.
Online courses (fast.ai, Coursera ML) are industry-recognised.

Tuition fees

Public degree RM1,500–RM4,000/yr; local Master's RM15,000–RM40,000.
The online route: RM500–RM3,000.

Scholarships

JPA/MARA/Khazanah; MDEC & national AI programmes offer subsidised training.

Certification & licence

No licence. None at all.
Unlike a Civil Engineer (BEM), an Accountant (MIA) or a Doctor (MMC), no body in Malaysia licenses software work. No register, no protected title, nobody who can stop you working.
What that means in practice:
A portfolio beats a certificate. Employers check the code you wrote, not the paper you bought. This is genuinely different from licensed professions, where the paper <i>is</i> the permission to work.
Be wary of an "IT diploma" sold as a shortcut. In a field with no regulator, nobody validates what that certificate is worth — and employers know it.
Bootcamps can work, but not because of the certificate — because they force you to build something. If it does not produce showable projects, it produces nothing.

Academic qualifications — stricter than other IT careers
Degree: Data Science, Statistics, Mathematics, Computer Science, Physics.
Master's: often required for a first post. This is the exception within IT — in most other IT careers a portfolio is enough. Here, the assumed statistical depth means academic qualifications still get checked.
PhD: needed for research work, not for ordinary industry work.

Legal duty
PDPA 2010 — models trained on customer data fall under Malaysian data protection law. Understanding this is part of the job, not an extra.

Certificates
Honestly, less important here than in most IT careers — academic qualifications and defensible projects carry more weight.
If you want one: a cloud machine learning certificate (AWS Machine Learning, Google Professional ML Engineer) is the most commonly seen.

What actually gets checked
Projects you can defend. Not "I ran this model on a Kaggle dataset" — but what the problem was, why you chose that approach, what failed, and how you know the result is valid.
Statistical judgement. Interviews test whether you understand what your model does, not whether you can call it.

Pros

Cons

Future & the AI era

LLMs are absorbing parts of classical modelling — but someone must evaluate, tune & govern those models: the DS.
The role evolves into "AI engineer + decision advisor". Your statistical grounding is an edge AI doesn't have.

Step by step, and how long each takes

  1. SPM — Add Maths matters hereSPM
    Unlike data analysis, Add Maths genuinely matters for this route.
    Data science is built on statistics and probability. A student who skips the mathematical foundations can learn to run a library, but cannot judge whether the result means anything — and that is the whole job.
    Concrete advice: if you are serious about this route, take Add Maths.
  2. Degree — statistics or computing3–4 tahun
    Data Science, Statistics, Mathematics, Computer Science, or Physics.
    Statistics and Mathematics are the best foundations for this route, though Computer Science is more common. The reason: programming is easier to teach yourself than statistics is.
  3. Master's — often required, and this is a real cost1–2 tahun
    This is the number people leave out. Most data scientist posts in Malaysia specify a master's, or take the candidate who has one over the candidate who does not.
    That means 1–2 additional years and additional cost before your first salary.
    Compare with a data analyst, who can start work straight after a degree. That difference — two years of earning versus two years of fees — deserves calculating before you choose.
  4. Statistics first, code secondKemahiran
    Python with pandas, scikit-learn, and one deep learning framework.
    SQL — you still have to get the data before you can model it.
    Statistics — and this is the part that separates people. Plenty can call a model. Far fewer can say whether the result is valid, whether the sample is biased, or whether the correlation means anything.
    That is the difference between a data scientist and someone running a library.

Key SPM subjects

Matematik Tambahan, Matematik, Fizik, Bahasa Inggeris

Add Maths →Matematik →Physics →English →

Questions students actually ask

Is data science really as hot as people say?
Partly yes, partly no — and the "no" half is rarely told to students.
What is true: the pay is good, the work is interesting, and demand exists in Malaysian banks, telcos and large e-commerce firms.
What is usually left out:
Entry-level posts are far fewer than in data analysis. Every company needs analysts; only companies large enough to have real prediction problems need data scientists. Plenty of data science graduates end up taking analyst posts — which is not a bad outcome, but is not what they were promised.
A master's is often required, so add 1–2 years and fees before your first salary.
The work is less glamorous than expected. A large share of the time goes on cleaning data, not building models.
Practical recommendation: do not abandon the field — but come in through data analysis. You are paid while learning, you find out whether you genuinely enjoy it, and you can move across with business experience a direct graduate does not have.
Do I need a master's to be a data scientist?
Usually yes — and this is the exception within IT, where most careers accept a portfolio in place of qualifications.
The reason is specific: data science rests on statistical judgement, and that is hard to prove through a portfolio. Anyone can show a notebook that runs a model. It is much harder to show that you know when a result is invalid.
When you can avoid it:
If you already hold a strong Statistics or Mathematics degree — that foundation is what the master's is expected to supply.
If you move across from data analysis with 2–3 years of experience and real production models. Experience can substitute for the qualification in this field, but it has to be the right experience.
Cost it honestly: a master's means 1–2 years without salary, plus fees. A data analyst over the same period earns RM3,500–5,500 a month and builds experience.
That is not a reason not to do it. It is a reason to decide deliberately rather than because it looks like the next step.
Practice SPM Add Maths →K1 papers + AI answers · 3 languages

Where to study — universities and total fees

University Course Total fees Duration Location
Multimedia University (MMU)Degree · BACHELOR OF COMPUTER SCIENCE(HONS) IN DATA SCIENCERM 62,2503 YearsCyberjaya, Selangor / Melaka
Raffles UniversityDegree · Bachelor in Data Science (Honours)RM 75,0003 YearsJohor Bahru, Johor
UCSI University / UCSI CollegeDegree · BACHELOR OF COMPUTER SCIENCE IN DATA SCIENCE WITH HONOURRM 76,2903 YearsCheras, Kuala Lumpur
Sunway CollegeDegree · Bachelor of Science (Honours) in Statistical Data ModellingRM 96,1203 Years (full-time)Bandar Sunway, Selangor
Sunway UniversityDegree · Bachelor of Science (Honours) in Statistical Data ModellingRM 107,4003 Years (full-time)Bandar Sunway, Selangor
Xiamen University MalaysiaDegree · Bachelor of Engineering in Data Science (Honours)RM 116,0004 yearsBandar Sunsuria, Sepang, Selangor
Swinburne University of Technology SarawakDegree · BACHELOR IN DATA SCIENCERM 121,9603 YearsKuching, Sarawak
Heriot-Watt University MalaysiaDegree · BSc (Hons) Statistical Data ScienceRM 131,7603 yearsPutrajaya
Monash University MalaysiaDegree · Bachelor of Computer Science in Data ScienceRM 136,8003 yearsBandar Sunway, Selangor
Monash University MalaysiaDegree · Bachelor of Applied Data ScienceRM 149,7603 yearsBandar Sunway, Selangor
University of Nottingham MalaysiaDegree · BSc (Hons) in Mathematics and Data ScienceRM 156,0003 YearsSemenyih, Selangor

Fees are a guide only and change every year. Confirm the current figure with the university before you decide.

Get all these colleges to contact me Tap to message Allite on WhatsApp. We pass your enquiry to the colleges above — no form to fill.

Sources

Figures and policy on this page are checked against these reports. Last reviewed 2026-09.

© 2026 Allite — Kerjaya selepas SPM