I’m a software engineer on the commodities desk at Qube Research & Technologies. I work on data systems, trading tools, and the interfaces that bring them together.
I’m studying Computer Science at NUS, focusing on parallel computing and database systems.
My work has taken me through trading systems, data engineering, and software for community organisations. Outside of work, I’ve helped build a patient management system for Project Sa’bai and contributed to Source Academy.
You can scroll through my experience and projects below, or explore them through the terminal.
More about meExplore from the terminal.
Type help to explore. Or try a command below.
Commodities desk.
Tech stack: Python, Pandas, NumPy, Airflow, AWS, AG Grid, Parquet, DuckDB.
Further details available upon request via email: yinjoeng2003@gmail.com
Email me for detailsShipped wallet features for TikTok LIVE while pioneering a file-level code-change impact analysis algorithm for its ~10M-line monorepo — surpassing tools like Nx — cutting redundant CI rebuilds and accelerating developer workflows by over 50%
Built a large-scale transcription and translation model with TypeScript, FastAPI, Redis, and K8s to ensure seamless compliance
Built high-frequency cryptocurrency trading systems in Python, C++, Kafka, ZeroMQ, TimescaleDB, and Redis
Improved the performance of liquidity monitoring, portfolio analysis, and performance analysis dashboards by over 20% through comprehensive bottleneck analysis and lazy loading
Engineered a workflow orchestration system to centralize management and monitoring of scheduled and ad-hoc view runs, boosting operational efficiency by 40% and providing real-time job visibility
Created an automated unit testing framework utilizing KDB+, Bash, AWS EC2, and GitHub Actions, eliminating dependence on external tools and saving significant licensing fees
Created data pipelines to extract, transform, and load real estate data from varied sources using Airflow, Kafka, and Python
Migrated the system codebase from legacy Ruby on Rails to React with TypeScript, creating reusable components across 42 functional areas, 150 pages, and over 300,000 lines of code
Architected a momentum-based real-time trade signal system processing over 500,000 equity data points per minute via WebSocket API, delivered to over 10,000 users
Headed a team of 12 software engineers in close collaboration with medical professionals to build a patient management system with Next.js, Django, and PostgreSQL, serving a user base of over 2,000 patients in Cambodia
Architected a face recognition system with AWS Lambda, DynamoDB, S3, and Rekognition which reduced registration time by 80%
Visit projectCollaborated with a professor in the CP3108A (Independent Work) course to customize the open-source JSCAD library and develop a hierarchical grouping system for 3D shapes in TypeScript, enhancing modeling efficiency by over 20%
Visit project