I built systems to improve the reliability & resource efficiency of distributed machine-learning pipelines, and reduced organizational friction slowing down experiments.

• Systematically profiled and optimized dozens of Apache Spark data pipelines via reducing scheduling contention, and tuning executor memory and concurrency. This increased cluster job throughput by 90% while reducing pipeline runtime by 30%
• Automated dataset lifecycle governance by integrating Hive databases into observability dashboards and notification workflows to identify and remove stale datasets. This reclaimed hundreds of petabytes of storage, reducing infrastructure costs by over USD 200k per quarter
• Led an organization-wide rollout of data caching tooling to reduce compute costs in data proprocessing stages during model training. This saved approximately 201k CPU cores in compute cost and improved GPU utilization rates by 8% on average
• Investigated and eliminated a prevalent Spark pipeline performance bottleneck by removing premature caching of large datasets, improving task throughput by up to >10× while reducing heap usage by up to 20%
• Diagnosed and fixed data quality issues in ML data pipelines, identifying incorrect downsampling, feature misalignment, and output anomalies affecting downstream model training
• Improved oncall and routine operational processes with regular team-wide technical sharing and writing of playbook guides to systematically solve typical oncall problems and handle day-to-day sysadmin operations

0
1

$ go run career.go
  compiling...

Espressif Systems develops wireless communication system-on-chips and cloud solutions to power IoT applications.

• Developed a chip power analyzer web application, using ReactJS to visualize data from an ESP32 devices and integrating a custom CNN model built with Keras to classify Bluetooth signals with 95% accuracy
• Assisted in conducting technical workshops, teaching undergraduates how to use advanced ESP32 chip features

The ToAI platform builds end-to-end IoT device management solutions for enterprise & government IoT needs.

• Developed Kubernetes cluster deployments to support a planned platform migration to use Kubernetes
• Used AngularJS to integrate enterprise APIs like PowerBI into the in-house IoT edge device management platform, enabling custom dashboard visualisations and wrote optimized SQL queries to generate geospatial reports for 32k+ devices

Testimonial & Award Nomination by my supervisor, Hsu Yee.

HackerTrail uses data-driven technology to evaluate and connect software professionals to employers.

• Spearheaded the migration of the candidate experience platform from Angular to React, rewriting key sections and implementing new features like interactive resizeable components to improve user experience
• Identified multiple high business value features missing on the platform via design thinking methodologies, contributing to the product’s roadmap development

Recommendation by my team lead on my SWE skills. (Scroll to recommendation section)

TOFFS provides Multi-CDN and cybersecurity solutions for Enterprise networking.

• Built concurrent data ingestion pipelines using Go, processing up to 2.5 TB/day of network telemetry data for machine learning workflows
• Reduced anti-DDoS model training time by 50% through GPU acceleration and optimization of training and preprocessing pipelines
• Automated model deployment and retraining workflows using Docker, Bash, and cron-based scheduling

Testimonial on my work by Co-Founder Jason Kong.

I was on the HPC team (Team Kent Ridge), stayed far too long on campus in a residential college (CAPT), and took care of community cats on campus 🐱

Complete coursework breakdown.

Awards & Honours

• GPA 4.7/5.0 (Highest Distinction), Deans’ List
  • Focus Area in Communications & Networking
• 3rd Place in ISC25 Online Student Cluster Competition as part of Team Kent Ridge
• Top Students, Software Engineering & Object-Oriented Programming (CS2113)
• Top Students, Engineering Principles and Practice 1 (CG1111A)
• NUS Computing Honour List of Student Tutors (awarded in 2025)
• CAPT Honour Roll (awarded in 2023)

Teaching

• Parallel & Concurrent Programming CS3211 ​, Spring 2025
• Real-Time Operating Systems CG2271 ​, Fall 2024
• Software Engineering & OOP CS2113 ​, Spring 2024
• Engineering Principles & Practice 1 CG1111A ​, Fall 2023 & Fall 2022