Data Scientist with 2+ years in banking — building fraud models that protect millions, and translating complex data into decisions executives can act on.
Technical Toolkit
Core tools and specializations I use day-to-day.
Work History
Building real impact in banking and beyond.
Selected Work
A mix of civic tech, open source, and medical AI.
A civic tech platform opening Thailand's national budget to public scrutiny — citizens can collaboratively review and upvote suspicious budget items.
Open-source Python library for Thai address processing: spelling correction and tokenization using Norvig, SymSpell, and edit-distance algorithms. Published on PyPI.
Trained FCN-8s, SegNet, and U-Net on 1,576 manually annotated CT scans (radiologist-verified) for hemorrhagic stroke segmentation. Delivered a clinical decision-support tool.
Academic Background
Built and evaluated CNNs (FCN-8s, SegNet, U-Net) on a manually annotated dataset of 1,576 CT scans, with annotations verified by a radiologist. Delivered a proof-of-concept tool to assist clinical interpretation.
Let's Connect
Whether it's a new opportunity, a collaboration, or just a chat about data — feel free to reach out.