Process 100+ daily financial transactions including deposits, withdrawals, drafts, bill payments, and cheque deposits with strict accuracy standards.
Verify client identification and apply appropriate hold policies following AML and KYC compliance guidelines for financial transactions.
Support account servicing requests including profile updates, PIN resets, and account maintenance while maintaining strict confidentiality standards.
Educate clients on digital banking tools and self-service platforms to reduce branch inquiries and improve operational efficiency.
Supported operations across five sold-out events at BC Place, Rogers Arena, and UBC Thunderbird ensuring smooth guest experiences.
Processed hundreds of POS transactions per shift while maintaining 100% cash accuracy during high-volume event operations.
Monitored crowd flow, safety compliance, and inventory levels to improve service efficiency and reduce service delays.
Mentored five new staff members on operational procedures, setup, and closing processes during high-pressure event shifts.
Completed 500–1,000 weekly annotations for LLM-based systems using specialized tools while maintaining strict quality and deadline requirements.
Evaluated AI model outputs for clarity, correctness, and alignment with instructions to improve machine learning training datasets.
Achieved 98–99% quality scores while resolving complex annotation edge cases to reduce downstream NLP model errors.
Escalated ambiguous responses and contributed to annotation guideline improvements reducing evaluator inconsistencies and project rework.
Resolved more than 300 weekly customer cases involving lost shipments, refunds, and product issues while maintaining 95%+ CSAT.
Investigated over 1,500 monthly orders using backend tools ensuring policy-compliant resolutions and reducing refund leakages.
Maintained top performance across AHT, quality metrics, and resolution accuracy while achieving strong SLA adherence.
Recognized as Employee of the Month in first month for exceptional performance and customer service quality.
Managed portfolio of 150+ high-risk customer accounts weekly using CRM systems and automated dialer platforms.
Ranked within top 20% of collectors based on recovery performance, compliance metrics, and quality audits.
Negotiated payment plans aligned with customer financial capacity while ensuring regulatory compliance and documentation accuracy.
Maintained accurate case documentation and account notes supporting regulatory standards and financial reporting requirements.
Developed multi-class weather image classification model using MobileNetV2 transfer learning architecture for high-accuracy image recognition tasks.
Achieved 91.69% test accuracy and 0.99 macro ROC-AUC through optimized model training and evaluation workflows.
Conducted exploratory analysis and feature evaluation to improve dataset quality and ensure model generalization across weather categories.
Implemented performance metrics including precision, ROC-AUC, and confusion matrices to validate model effectiveness and prediction reliability.
Designed a metadata-only multi-class music genre classification model using XGBoost to analyze structured music metadata features.
Achieved 83% test accuracy with a 0.83 macro F1-score and 0.98 ROC-AUC across multiple music genre categories.
Performed feature engineering and feature importance analysis to identify the most predictive metadata attributes.
Evaluated model performance using cross-validation, classification metrics, and confusion matrices to ensure reliability and predictive stability.
Designed and published an interactive Tableau dashboard analyzing Canadian COVID-19 case-level data across provinces and demographic groups.
Developed geospatial visualizations highlighting regional hotspots and high-risk demographic segments across Canada.
Applied data cleaning and transformation techniques in Excel to prepare datasets for dashboard integration.
Enabled faster public health insights by presenting interactive charts, filters, and data storytelling elements for decision-makers.
Processed and analyzed dataset containing 10,000 Spotify tracks to identify patterns influencing music popularity predictions.
Implemented ensemble machine learning models including Voting Classifier, Balanced Random Forest with SMOTE, and XGBoost.
Achieved 0.82 macro F1-score and 0.90 ROC-AUC improving prediction reliability compared to baseline dummy classifiers.
Performed exploratory data analysis, feature engineering, and model evaluation using classification metrics and visualization techniques.
Developed a full-stack reservation management system automating booking operations using PHP, MySQL, and React.
Implemented CRUD functionalities enabling real-time booking creation, updates, deletion, and centralized management of reservations.
Reduced booking errors by 35% by designing structured data storage and validation mechanisms within relational database systems.
Enabled efficient reporting and operational insights through structured relational datasets generated from booking records.
Open to data analytics, business intelligence, and data science opportunities focused on data-driven insights and reporting automation.
Copyright © Jae-Hyun Park 2025. All rights reserved.