The household sector balance sheet provides insights on the assets and liabilities held by households in Singapore as at end of period. The current compilation workflow involves manual data extraction from multiple different sources with varying input formats and data frequency, and tedious updating of charts and tables for reporting purposes.
This project aims to overhaul the end-to-end workflow as follows: i) standardising the ingestion of diverse data sources, ii) migrating the compilation process from FAME and manual handling to a Python-based workstream, and iii) revamping the way compiled data is presented for submission and reporting, including exploring the integration of AI for data analysis and dashboard generation.
iv) introducing automated data validation checks at the point of ingestion, e.g. outlier detection, and completeness checks. v) Building a centralised metadata repository, e.g. capturing data definitions, update frequencies, source contacts. vi) Automating production of narrative reports or commentary templates vii) Documenting new workflow and running structured handover sessions for all users viii) Preparing presentation deck on new workflow.
Through this project, we will develop a standardised data ingestion framework, a Python-based compilation workflow to replace existing FAME-dependent processes, and an automated reporting output or dashboard for timely data reporting. In doing so, the selected intern will gain hands-on experience in statistical data processing and compilation, and develop practical skills in Python and AI applications.
Proficiency in Python and R programming Knowledge or experience in computer science, data science, data analytics, programming, data processing, or process automation. Internship Period: 6 months, from Jan to Jun 2027 (flexible start and end dates) Part Time option available, min.
3 days per week