Making the 'Digital Twin' of global food system more accurate using machine learning

● Solved client bugs and developed workflows for adsorption in carbon capture and mineral extraction.● Created a VSC plug-in to provide highlights and hover descriptions for enhanced user experience.

Predictive r-adaptation using Deep Learning● Investigated machine learning-based methods to accelerate mesh adaptation for solving partial differential equations, for applications like ocean modeling, astrophysics, and bioengineering● Developed a graph neural network (GNN) model to replace the Monge-Ampère equation in mesh adaptation with an accuracy of 95% in movement pattern prediction

● Assisted in developing and implementing a data-driven process flow optimization strategy that reduced the average duration of the product development process by 8% ● Coordinated with cross-functional teams to redesign information flow routes and optimize collaboration, resulting in improved efficiency and quality of work

- Reduced energy usage by 5% and consolidated process flow data with a Sankey Diagram for the overall system- Revised thermal and mass balance framework for insulator line to reduce the production time