Responsibilities include:

• Utilizing Python’s pandas and sklearn libraries to efficiently process large volumes of non-imaging data from various cohorts to create a master dataset for training the model
• Working on developing an end to end transformer based pipeline for identifying different dementia etiologies using brain scan MRI and non-imaging data
• Researching ways to improve the model performance by experimenting with different loss functions and utilizing different strategies to tackle the class imbalance problem

Responsibilities include:

• Built a reliable machine learning framework using BERT backbone to recognize the semantic difference between mentions of race vs. mentions of color in non-racial terms in the media articles
• Extracted racial keywords from a sentence using attention analysis of the trained model’s hidden layers

Responsibilities include:

• Worked on development and testing of embedded Linux Flex operating system
• Automated the process of embedded testing using Unix test scripts and CI/CD tools like Jenkins and LAVA