Design end-to-end machine learning architectures that support geospatial analytics and demand forecasting use cases, including the organisation's Spatial Modelling Engine.
Define robust, maintainable, and extensible data pipelines, feature engineering strategies, and model-serving frameworks.
Ensure that architectural decisions account for the long-term nature of infrastructure planning, where model outputs must remain interpretable and auditable over multi-year periods.
Machine Learning Development and Implementation
Develop, test, and deploy machine learning models and geospatial analytics solutions within a production environment.
Build and maintain data pipelines that integrate a variety of data sources, including housing development information, demographic records, migration patterns, land-use plans, and accessibility metrics.
Collaborate with engineering and platform teams to ensure that models are reliably operationalised and continuously monitored over time.
Machine Learning Optimisation and Geospatial Analytics
Develop and enhance predictive and spatial models that forecast future education demand across Singapore's planning landscape.
Apply suitable techniques based on the problem context, including spatial regression, time-series forecasting, agent-based modelling, and deep learning.
Continuously assess model performance, validate outputs against ground-truth data, and refine modelling approaches to improve forecast accuracy and reliability.
Qualifications
At least 3–5 years of hands-on experience in data science or a related field, with a demonstrated track record of delivering machine learning solutions in production.
Prior experience working with geospatial data and tools is strongly preferred.
Experience in areas involving demographic modelling, urban planning, or public-sector analytics would also be highly relevant.
Familiarity with Singapore's planning context, including URA Master Plan data, HDB housing pipelines, or similar datasets, would be an advantage.
Required Skills
Proficiency in Python and relevant data science libraries, such as scikit-learn, PyTorch, or TensorFlow.
Knowledge of geospatial tools and frameworks, such as GeoPandas, QGIS, PostGIS, or ArcGIS, would be an added advantage.
Strong SQL skills and experience working with cloud data platforms, such as AWS, GCP, or Azure.
Familiarity with the full machine learning lifecycle, including data wrangling, feature engineering, model evaluation, deployment, and monitoring.
Comfortable applying advanced machine learning techniques, such as ensemble learning, regularisation, agent-based modelling, and forecasting.
Strong communication skills, with the ability to clearly present findings and recommendations to non-technical stakeholders.
Ability to collaborate effectively within a cross-functional team environment.
