About
Honours student at USyd researching neural network stability and system reliability. I bridge the gap between rigorous mathematical physics and production-grade software—building scalable PyTorch pipelines, executing large-scale HPC workloads, and processing high-throughput time-series and document datasets.
Professional Experience
Skills & Toolkit
Check out my latest work
Currently working on a variety of projects, from simple tools to full-stack pipelines. More to come soon!
High-throughput command-line engine that converts complex, unstructured PDFs into RAG-optimized Markdown using Vision-Language Models. Built with thread-pool concurrency, automated API backoff retries, and stack-based post-processing to repair broken LaTeX notation and page-split boundaries.
A selection of travel, wildlife, and landscape photography. Captured on my DSLR and color-graded in Lightroom.
Get in Touch
Have a role or collaboration in mind? Feel free to reach out directly at contact@yihao.au or connect via LinkedIn.