Working on RAG, AI performance evaluation and benchmarking, and core services.

• Worked at the Psychiatry Neuroimaging Laboratory at Brigham and Women’s Hospital, Harvard Medical School

• 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

• Built a data pipeline using AWS Lambda and Pandas to collect and process logs from semiconductor chips under different configurations, reducing manual set-up time by 50%

• 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

• Managed project goals and tasks among 7 developers

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