Assist people living with Dementia through AI/ML, via an interactive game called DevaWorld. Tasked with the goal of developing and training an ML model to predict the players' actions/engagement in order to spot any medical emergencies without the full reliance of a physical caretaker. Extracted, cleaned, classified and validated a large video dataset. Used metrics such as Fleiss’ kappa to measure the consistency of the data labeling guidelines that I revised. Tested various ML models with a CNN-RNN architecture, using TensorFlow, Keras, and pandas on a subset of the training dataset. Compared their performance and accuracies which devised which model to train the data.

Specialized in evaluating and optimizing LLMs through systematic testing and analysis of code and mathematical prompts. Provided detailed qualitative reviews to improve model performance on code generation and mathematical reasoning tasks. Contributed to model refinement by identifying edge cases and areas for improvement in technical prompt handling.

The Career for Access to Engineering Excellence "CAEE" Undergraduate Tutor. Taught in subjects such as linear algebra, multivariable calculus, data structures, computer programs, data science.