May 2024 - Present
Technical Desk-Align
- Designed and maintained in-house Python packages, automated jobs, and API endpoints, collaborating effectively with a global team of over 15 developers
- Took on roles of sprint manager and support personnel to address queries and issues from Quant Researchers
Jul 2022 - May 2024
Enterprise Risk
- Enhanced departments' capability of computing risk, performance, and exposure metrics for top-of-the-house reporting by developing in-house Python packages, computation engines, and backend APIs
- Drove project progress by orchestrating coordination between developers and end users, gathering requirements from end users and scoping the work, ensuring timely delivery ahead of milestones
- Ideated and developed open-source Python package for working with tree data structures, with over 100 GitHub stars to date
Aug 2018 - Jul 2022
Part of the Business Analytics Graduate Immersion Programme (BAGIP) which is a 12-month accelerated programme focused on building core analytical skills essential to business growth
CBO Call Centre Call Reduction Project (CCTR)
Time Series Forecast Reusable Asset (FORA)
CBO Cards MTJ Project
CBG Mortgage Project
Audit Summarization Project
T&M FX Project
Feature Mart Reusable Asset
Other Projects
Others
CBO Call Centre Call Reduction Project (CCTR)
- Built, deployed, and monitored ML pipeline and performed system integration testing to predict and address customers' needs prior to call
- Initiated and single-handedly improved data and feature engineering pipeline architecture to standardize processing, reduce resource wastage, and increase scalability, resulting in time savings of over 35 hours
- Delivered data analysis and insights to various stakeholders in accordance to business requirements to track model performance, support experimentation, and drive decision making
- Performed predictive analysis with rule-based model as a real-time use case using Adobe IBMB clickstream data
- Published in IJCNN 2020 Personalized Digital Customer Services for Consumer Banking Call Centre using Neural Networks
Time Series Forecast Reusable Asset (FORA)
- Developed forecasting methods for time series, involving EDA, processing, models, and pipelines, subsequently packaged and deployed it to DBS internal server using Jenkins
- Implemented end-to-end forecasting and anomaly detection workflows with CBG, IBG teams across SG, HK teams to forecast trade volume, CASA balances and transactions, metrics related to cards onboarding and more
- Held workshops, demonstrations, project presentations, and code walkthroughs to business and technical teams from SG, CN, TW, HK and ID
- Independently created and launched an enterprise-wide web application encompassing time series forecasting methods using Dash and Plotly, and deployed it using Jenkins and OpenShift
- Launched 2 forecasting e-learning courses for bank-wide staff, collaborating with Culture and Curriculum team, and have over 1500 user completions to date
- Voted top Reusable Asset (RA) in RA Learning Festival 2020
CBO Cards MTJ Project
- Developed end-to-end anomaly detection and forecasting pipeline to predict metrics related to card onboarding journey
- Deployed solution on QlikView dashboard within 5 months of project inception, making it the first AI-embedded Control Tower within the bank
- Published in ACM 2022 Improving Operational Efficiency through Predicting Credit Card Application Turnaround Time with Index-based Encoding
CBG Mortgage Project
- Developed end-to-end ML solution for mortgage pricing to predict customers' reprice rate
- Improved existing data pipeline architecture, resulting in 85% time savings
- Managed project deliverables, deployment plans, and drove project to deployment in 6 months
Audit Summarization Project
- Performed ideation, experimentation, evaluation, and benchmarking of different models for NLP summarization task for Audit issues
- Tested and deployed API solution on GPU in UAT
T&M FX Project
- Developed pipelines to automate campaign analysis, standardizing processing across 2 campaigns and 3 types of campaign analysis
Feature Mart Reusable Asset
- Formalized and implemented Feature Mart which automates feature creation for CBG data and standardizes processing required
Other Projects
- Performed email clustering using different text cleaning and embedding methods and machine learning models
- Implemented best practices workflow in multiple projects by setting up integration testing of codes and enabling distributed processing
- Have experience with Kedro-style pipelines
Others
- Spot Award recipient for 2021 Feb, Jun, Oct, and 2022 Apr for exemplary work performance
- Star Award recipient for first half of 2021 for exemplary work performance
- Second most active Coursera user, based on course completion, in the company in 2018
- Completed over 100 Coursera courses within the span of 3.5 years
- Represented Musicians Interest Group in DBS Dinner and Dance 2019 and 2021, playing keyboard in a live band
May 2017 - Jul 2017
- Review of IT processes and controls over logical and physical access, password management, user account management, audit logging, program change management, system development lifecycle, computer operations etc., governing data integrity, confidentiality and availability
- Attended professional training courses in assurance reporting and SAP modules
- Exposed to banking risk management, custody, data analytics, global financial services, prime brokerage, private equity, securitization, treasury and various frameworks
- Represented Topaz (Team Advisory) in KPMG Sports Carnival 2017, clinching first place in Cheerleading and second place in Touch Rugby, assisting Topaz to secure Overall Champion
Aug 2016 - Aug 2017
- Assist in building networks within campus and with the employer brand and recruitment team at Deutsche Bank
- Influence and shape Deutsche Bank's on-campus marketing activities and recruitment strategies