Proof-of-concept ML projects for product demand forecasting and tractor fuel consumption prediction — with modern delivery tooling.
Stakeholders needed quick, credible forecasts for product demand and equipment fuel use — packaged as usable POCs rather than notebooks alone.
Implemented two forecasting POCs and gained hands-on experience packaging ML apps with Docker, Streamlit UIs, FastAPI services, and Azure AI capabilities.
Established a foundation for production ML delivery patterns later applied to pharmaceutical and inventory systems as Associate Software Engineer.