Policy Researcher (Data Science)

About Khazanah Research Institute

Khazanah Research Institute (KRI) was formed in 2014 and is funded by Yayasan Hasanah and Khazanah Nasional Berhad. KRI is independently governed by its own Board of Trustees and provides independent views that enrich public policy research and advocacy space in Malaysia.

KRI is dedicated to Advancing Malaysia through rigorous, evidence-based research that results in actionable policy recommendations that improve the welfare of Malaysian families, uplift their communities and drive inclusive progress.

For more information on KRI, visit www.KRInstitute.org and follow us on social media, @KRInstitute on X, Facebook, Instagram, TikTok and Khazanah Research Institute on LinkedIn and YouTube.

About the Role

KRI is recruiting researchers whose primary strength is computational: people who build the datasets, pipelines, models and tools that let policy questions be answered well, and who can explain what the results mean to the people who make decisions. You will work across KRI's research areas, partnering with policy researchers on projects where the data is large, messy, spatial or unstructured, or where the question calls for simulation, modelling or machine learning rather than conventional analysis. Recent examples of this kind of work at KRI include building a Greater Kuala Lumpur mobility dataset from transit feeds and field observation .

We welcome applications from fresh graduates through to mid-career and senior researchers for the roles of Research Associate/Senior Research Associate. The level of appointment is determined by KRI on the basis of experience and demonstrated past work.

If you are early in your career, we are interested in what you can build and how you think. Please read “To Apply” before deciding whether you qualify. Coursework and student projects count as evidence, and we do not expect you to arrive with an established public record.

What You Will Do

• Build and maintain reproducible data pipelines from survey, administrative, geospatial, transactional and web sources, documented so that a colleague can run them and obtain the same result.
• Develop models to answer policy questions: statistical, causal, spatial, simulation or agent-based, and machine learning where it fits.
• Work with policy researchers to turn a policy question into a tractable empirical design, and turn results into research papers, policy briefs, short articles and data products.
• Publish datasets, code and interactive outputs that others can use and verify, and build on. Work in the open by default, in version control, with the analysis documented well enough that a reader can reproduce the result independently from raw data.
• Engage policymakers, regulators, civil society, the media and the public through briefings, public commentaries and op-eds, and public events, explaining technical results in plain language.
• At senior level, set technical standards and develop the computational skills of other researchers.

Who We are Looking For

• A degree in computer science, applied mathematics, statistics, physics, data science, economics, or any discipline in which you have built quantitative models or data systems to a high standard.
• Code in Python and R that another person can read, run and check, kept under version control; SQL for working with databases; Stata, in addition, is useful.
• The ability to take a raw, inconsistent dataset and make it analysis-ready, and to explain the judgment calls you made.
• The judgment to choose a method that fits the question, and to say plainly what it can and cannot show.
• Evidence of real work: for recent graduates, a substantial project (thesis, capstone, competition entry, open-source contribution or work project) that you built yourself and can walk through in detail; for experienced candidates, a data product or model you have owned end-to-end, and evidence that your work informed a decision.
• Clear written communication: able to explain a technical result to a non-technical reader in two pages.
• Interest in Malaysian public policy. Prior policy experience is not required; familiarity with Malaysia's socio-economic context and data sources (e.g. Department of Statistics Malaysia (DOSM) surveys) is an advantage.

An Advantage

• A Master’s or PhD.
• Experience with at least one of: geospatial data, network data, simulation or agent-based models, microsimulation, text or web-scraped data, or large administrative records.
• Familiarity with Malaysian data sources, for example, DOSM’s household surveys. For experienced candidates, a data product or model you have owned end-to-end, and evidence that others have used it.

What We Offer

• Institutionally funded research: Focus on research without the burden of grant applications.
• Autonomy and impact: Pursue research aligned with your interest or specialisation and shape policy discourse.
• Professional growth: Training, conferences, and opportunities to publish and present your work.
• Cross-sector exposure: Direct engagement with government, academia, civil society, and international research institutes.
• Flexible working: A hybrid working arrangement and a culture built on mission, evidence and public purpose.

To Apply

Email [email protected] with the subject line: Application for Researcher (Data Science): [Your Name] .

Please include:

• CV
• A cover letter (one page) on what you have built, which Malaysian policy problems you would want to bring it to, and why KRI.
• Academic transcripts and certificates.
• At least two sole-authored writing samples (published or unpublished). You may include additional co-authored work; if so, please state your own contribution.
• A portfolio of code you have written and can walk through in detail. This may be a public repository, a final-year project or dissertation, coursework, a competition entry, an open-source contribution, or work from a job, internship, or student organisation. What matters is that you built it, you can explain it, and you can share it with us. We are not looking for an established public record, and we do not expect early-career applicants to have one. We do expect you to be willing to work in the open. If everything you have built is proprietary or otherwise cannot be shared, submit instead a small piece of work written for this application, in a public repository, using public Malaysian data. Describe two of your proprietary projects separately (what you built, the tools, your role and the outcome) and be ready to walk through one at the interview.

Shortlisted candidates will be asked to complete a short, bounded data task using public Malaysian data, or, for experienced candidates, to present a piece of their own work. The task is designed to take about four hours, and the deliverable is capped so that it cannot reward spending longer. We assess how you handle messy data, the judgment calls you make, and how clearly you explain them.

Applications are reviewed on a rolling basis. We regret that incomplete applications cannot be processed. Due to the volume of applications received, only shortlisted candidates will be contacted. Thank you for your understanding.