Building AI that actually solves real-world problems.

Co-built and load-tested the backend for a university exam scheduling system handling room allocation and invigilator assignments, deployed on AWS with auto-scaling (1–4 instances) achieving 99.9% uptime.

Implemented and fine-tuned a Siamese network to classify the ISIC 2020 Kaggle Challenge dataset and achieved 80% accuracy despite a severe 98:2 class imbalance.

Contributed to building a LangGraph-based sales automation agent at useoven.com (UQ ilab accelerator), reducing manual outreach tasks by 90%.

Collaborated in a team of 4 to develop and deploy an internal Flutter app for real estate listings; improved app performance by 30% through UI component optimisation and asynchronous data handling in Dart.

Facilitated tutorials and practical sessions (4 practicals and 1 tutorial) for second-year CS students in Theory of Computing (COMP2048).