At a NdFeB magnet manufacturer, I report to the Deputy CEO on a greenfield plant, track equipment specifications and FEED tasks, and develop price benchmarks from Chinese market data and company annual reports. I also supported CAPEX and OPEX estimates, DPR sections, and bid submission.
On Rare Earth Intel, I built a Python pipeline that extracted 877 figures from 26 PDF reports with page-level traceability, then validated the data against ground truth. I added question answering restricted to verified data and tested it with pytest. My interest in machine learning began with an academic autonomous vehicle project and grew through project work in the magnet industry.
