Data & AI Engineer
Hey
I'm a Research Assistant @ UOW, currently researching how to keep the data behind RAG systems accurate as documents change. I graduated in Computer Science with a Distinction and love building things with AI APIs like Gemini and DeepSeek.
Let's TalkA private Telegram bot I use every day to log my cats' spending, weight, flea meds and vet visits. I can ask it questions in plain English. An LLM works out what I mean, but the totals come from SQL, so it can't make numbers up.
Open-source CLI that turns PDFs into Markdown, JSON and chunks ready for embedding. It also pulls text out of images, checks the OCR quality and falls back to Gemini vision when OCR does badly. It has 26 pytest tests.
A RAG pipeline that knows which version of a document is current. Plain vector search kept returning outdated versions, so I added version filtering per question. It only re-processes files whose SHA-256 hash has changed.
Cleaned 20 years of Hong Kong Observatory data. The annual mean temperature rose about 1°C between 2006 and 2025, faster than the global average.
Sep 2026 - Present
University of Wollongong
I'm helping write a paper on keeping an AI assistant's knowledge base accurate as documents change. I found that, by default, text inside PDF images never makes it into Docling's chunks. I compared four OCR approaches and designed a fallback to Gemini.
Jun 2019 - Dec 2024
Amalfitana / Fratelli · Repulse Bay
Promoted from Bar Supervisor. I ran the bar team at a busy restaurant for three years.
Jan - Apr 2019
HKEVPOWER · Science Park
Supported EV charging station installation and field operations. Reference letter >>
2026
University of Wollongong
Distinction average · 78% WAM
2019
IVE
GPA 3.4 · ISTS Innovation Award
Toolkit
SQL (PostgreSQL) · Python · pandas · Power BI · DAX · Docling · pgvector · Ollama · n8n · AWS · Git
I'm looking for junior data and AI roles in Hong Kong. The quickest way to reach me is email.
Send Me A Message