About
Hongda, like Honda. Yes, the car.

I was born in Harbin, went to middle and high school in Singapore, and came to Canada for university: a first Computer Science degree at the University of Toronto, then an Honours BSc in Computer Science at York’s Lassonde School of Engineering.
Today I’m an AI Agent Developer at Geotab in Toronto. I was the first engineer on the platform and, for most of its life, the only one. I designed the architecture, wrote the code, and run it in production. It is a multi‑agent system on Google ADK and Vertex AI: a root agent, a document sub‑agent wrapped in deterministic validators, retrieval with citations over a governed corpus, typed tools into BigQuery, Google Maps and the HR system, and self‑hosted Langfuse tracing, because the data cannot leave company infrastructure. The corpus is governed as code: a manifest reviewed in merge requests, and a daily job that watches the sources for drift.
Before that I spent a year on Geotab’s data enablement team building Airflow pipelines, DataHub lineage and Superset dashboards, and a summer at the Ontario Ministry of Health moving datasets from Oracle to Redshift. In 2023 I worked with Nick Feamster on email classification; the paper is in SPIE proceedings.
The one line people use about me is “he gets things done.” I don’t mind challenging the status quo when it improves the outcome.
Off the clock I play the sanxian, a three‑stringed Chinese lute. I’ve performed at the Toronto Dragon Festival, with the Toronto Artist Symphony Orchestra, and for the Chinese Embassy in Canada, which gave me an Outstanding Cultural Contribution award. I train jiu‑jitsu four times a week and have played badminton since high school.
Experience
Résumé (PDF)- 2026 –AI Agent Developer, Geotab
First engineer on the team and owner of the agent platform end to end: architecture, code, production. Wrote the technical screen and hired the second engineer. Named the division’s AI technical expert; AI Champion award, 2026.
- 2024 – 25Data Developer (intern), Geotab
Airflow pipelines across production and UAT, metadata and lineage on DataHub via the Python SDK and REST API, Superset dashboards. Worked with the Embedded, RevOps, Keyless and Data Platform teams. First agent work, on Google ADK.
- 2024Application Programmer, Ontario Ministry of Health
Business intelligence team. Automated the migration of 30+ datasets from Oracle to Amazon Redshift with Python and S3, built 12 IBM Cognos dashboards, and wrote a pandas script that halved the time colleagues spent comparing Excel reports.
- 2023Machine learning research, with Nick Feamster
Email classification on a 5,172-message dataset. The improved SVM reached 98.17% accuracy; published in SPIE proceedings (ICCAID 2022).
Education
- 2025Honours BSc, Computer Science
York University, Lassonde School of Engineering
- BSc, Computer Science
University of Toronto
Geotab, in three pictures
More photos


Elsewhere
- GitHub
- Calendar
This site is Astro on Cloudflare, set in Inter and JetBrains Mono. No analytics, no cookies.