Implemented an implicit neural network using a grid-based hash-encoding architecture, resulting in a 3x decrease in training and inference times on ultra-high resolution MRI scans

Assessed performance using established metrics, achieving a 10% boost in image reconstruction quality

Developed a random configuration selection logic using AWS Parameter Store, reducing 90% of time bias in data

Built an interactive data visualization dashboard using Dash Plotly to report insights to various stakeholders

Developed an internal applicants screening tool, reducing screening time significantly (NextJs, FastAPI, Firebase)