At Permodalan Nasional Berhad, I designed and deployed a Python-based validation framework for financial datasets with 100k+ records, reducing manual review time by 99.996%. I also built reconciliation pipelines to detect inconsistencies in private market transaction records.
I led system-level UAT, checking dataset integrity and investigating processing failures. I generated analytical reports summarizing validation findings and risk exposures.
At KENANGA INVESTMENT BANK BERHAD, I built financial forecasting and valuation models using Excel and Python, and automated regulatory compliance checks, reducing document review time by 99.7%. I also extracted and structured financial data from Bloomberg and regulatory filings.
At KHAZANAH NASIONAL BERHAD, I developed a framework combining FinBERT sentiment analysis, factor modelling, and machine learning to forecast US equity returns. My projects have also included LSTM-based stock return prediction, portfolio backtesting, and statistical analysis for a national datathon.
