About
Hey, I'm Toby.
I studied New Media at City University of Hong Kong's School of Creative Media, where code, design, and interaction meet. Somewhere between generative art and HCI projects I fell into Data Science, and discovered I'm passionate about turning data into insights that solve real problems.
I co-founded HKLawyer.ai because I saw firsthand how inaccessible legal help is for regular people (like me). I own the whole stack, from the hybrid-search RAG pipeline (dense + BM25 over 20+ years of judiciary data) to the frontend and deployment. It has answered 20,000+ questions for 52K+ visitors, and an LLM-as-a-judge benchmark keeps it honest.
I'm practical about Data & AI. I care deeply about understanding the data, not just the models. Whether it's analyzing user behavior patterns, building predictive models from scratch, optimizing ETL pipelines, or creating dashboards that tell a story, I dive in.
Most recently I was an Associate AI Engineer at DHL Data & AI in Bonn, building Document AI pipelines on Gemini, a Pydantic AI multi-agent system, and GenAI tools. Before that I spent almost five years as a Data Analyst at ALTEN in Hamburg: Power BI for 2,000+ consultants, Fabric and Databricks pipelines, and my first production RAG systems. Along the way: BASF and the Hong Kong Census and Statistics Department.
My M.Sc. thesis at Leuphana asks how to stop LLM agents from drifting on long tasks: a deterministic engine grounds every decision in computed facts, and evolutionary prompt optimization does the rest. I'm now open to AI engineering, research software, and applied research roles.
I speak Cantonese natively, Mandarin and English fluently, German at B1, and Python every day. I live in that sweet spot between data analysis, machine learning, and making insights actually usable.
My creative media background? It's why I obsess over clean visualizations and making complex data feel intuitive, because insights should be accessible, not buried in spreadsheets.
More: CV · experience · skills · uses · say hi

visitor@thetobysiu:~$ cat philosophy.txt
Good design is obvious.
Great design is transparent.
— Me, probably inspired by someone smarter
visitor@thetobysiu:~$ echo "AI Engineer"
class KingWaiSiu: def __init__(self): self.role = "AI Engineer & Data Analyst" self.location = ["Bonn", "HK"] self.expertise = [ "Data Analysis", "LLM Integration", "RAG Systems", "Full-Stack AI", ] self.current = ["HKLawyer.ai", "open to new roles"] def build_cool_stuff(self): return "Always." # no exceptions
def reach_out(): email = "thetobysiu@gmail.com" location = "Bonn / Hong Kong" available = True interests = ["AI Projects", "Collaborations", "Coffee Chats", "Opportunities"] if you.have_cool_idea: send_email(to=email) return email