Early-career Data Scientist, 0 to 2(YOE)
You build the models behind credit and risk decisions, for businesses and for people. Apply by doing a short take-home first; the steps are at the end of this post.
Link to the assessment: *************
The Company
OneCredit is Malaysia's newest credit bureau, registered under the Ministry of Finance. We turn credit and business data into insights that help businesses and consumers manage risk and make informed decisions.
The Role
We're hiring a Data Scientist to build the models behind our customers' credit and risk decisions. You'll work with product managers, ML engineers and other data scientists to explore data, build features, train and validate models, and explain the results so they hold up in a regulated setting. Hands-on work, plenty to learn, real ownership early.
What You'll Do
- Explore new datasets and check their quality
- Build and document features from filings, financial statements and court records
- Build training and evaluation sets that keep the future out of the past
- Train and compare models, from simple baselines to gradient boosting, NLP and LLMs
- Evaluate properly: the right metrics, out-of-time testing, error analysis
- Work with ML engineers to ship models and track them in production
- Write documentation and present your findings
What We're Looking For
- 0 to 2 years in data science, analytics or a related role. Fresh graduates with relevant projects or internships are very welcome
- A degree in statistics, maths, economics, computer science, actuarial science or similar
- Solid statistics, hands-on ML in Python, and good SQL
- You understand leakage, class imbalance, and why accuracy is often the wrong metric
- Clear communication in English
- Plus: credit risk, NLP or LLM projects, entity matching, financial statements, fintech
Tech Stack
Python, SQL, Snowflake, dbt, Apache Airflow, scikit-learn, XGBoost/LightGBM, SHAP, MLflow, LLM APIs, AWS, Datadog, Langfuse.
Why OneCredit?
- Build models that shape real credit and risk decisions
- Credit risk, NLP and LLMs in one team
- Ownership early, with mentoring and review on everything you ship
- Early team, fast progression for high performers
Kuala Lumpur, 3 days in office.
Interview Process
- Take-home → HR screen → technical interview (a walkthrough of your take-home) → CEO
- To apply: do the take-home at ************* (6 to 8 hours), then send it with your résumé through the form linked in its README.
- Everyone who submits gets written feedback by email.