Want to know if this job is worth applying to?
Kuala Lumpur, Malaysia
Onsite
Job Description
As a Data Scientist at AirAsia, you won't just be crunching numbers; you’ll be building the engines that power Asia's leading travel and fintech platform. You will collaborate with cross-functional teams to solve complex problems in revenue management, personalized marketing, and operational efficiency. We are looking for someone who loves the "loss function" as much as the "user experience."
Model Development: Design and deploy predictive models (classification, regression, and clustering) to optimize flight pricing, ancillary sales, and customer churn.
Experimentation: Design and analyze A/B tests to validate product features and marketing campaigns, ensuring results are statistically sound.
MLOps & Engineering: Collaborate with Data Engineers to productionalize models. You will help maintain the ML lifecycle, including versioning, monitoring model drift, and automating retraining loops.
Insight Generation: Translate complex statistical findings into "Allstar-friendly" insights for stakeholders and executive leadership.
Feature Engineering: Architect high-signal features from raw clickstream, booking, and aircraft sensor data.
Education:
Bachelor’s or Master’s in Data Science, Computer Science, Statistics, or a related quantitative field.
Other Requirements:
Machine Learning: Strong foundation in supervised and unsupervised learning. Experience with frameworks like Scikit-learn, TensorFlow, or PyTorch.
Cloud Knowledge: Proficiency with Google Cloud Platform (GCP)—specifically BigQuery for data extraction and Vertex AI for model orchestration.
MLOps: Familiarity with CI/CD pipelines, Git version control, and containerization (Docker/Kubernetes) to ensure models are scalable and reliable.
Analytics: Expert-level SQL and Python. Experience with A/B testing methodologies (hypothesis testing, p-values, and confidence intervals).
Experience with Large Language Models (LLMs) for feature augmentation or chatbots.
.png?1790775966){kind=logo}