Data scientist & AI engineer building time-series models, data pipelines and reliable backends — currently a Master's candidate in Data Science & AI at the Asian Institute of Technology.
I study how messy, real-world data turns into decisions — and I build the pipelines, models and small tools that make that crossing reliable.
Originally from Myanmar, now in Thailand, I've followed a slightly unusual academic route — four institutions across three countries, in parallel — and a working life built around forecasting markets and shipping production systems remotely.
At Gold Silver Central I led the AI team designing ARIMA-based forecasting for gold markets, building end-to-end pipelines on Azure and trimming a third of our recurring manual work via APIs and automated reports.
Outside paid work, I teach data science to people just getting started — which is, honestly, where most of the learning happens.
ARIMA time-series modelling of gold spot prices combined with technical indicators. End-to-end ingest → clean → transform pipeline, served via API + WebSockets on Azure, with Locust-backed reliability testing.
A maritime monitoring tool: software framework plus GPS, depth meter and wind indicator modules, backed by an SQLite store. Built agile, demoed to clients with Canva-based mood boards before each iteration.
A DataCamp workspace project — segmenting sales channels, evaluating performance with statistical tests, and recommending a revenue-maximising distribution mix.
Group capstone for MIT Emerging Talent — cleaning & reconciling cross-national literacy datasets, then surfacing actionable signals for educational interventions in developing nations.
Front-end for a computer-vision diagnostic tool — clinicians upload a chest X-ray and receive a model-driven COVID-19 classification with confidence signals.
