Harini is a PhD student studying applications of machine learning at MIT. She also completed her B.Sc. and M.Eng. in EECS at MIT previously. In the past, she has worked on ML for healthcare, demonstrating effective and interpretable methods for predicting onset and weaning of invasive interventions for patients in Intensive Care Units. Her other projects include studying language disparities across economic class and developing medical image segmentation methods for malaria detection. Currently, she examines the unintended
consequences of automated decision-making in real-life scenarios with flawed data.
Harini has done several internships at tech companies. Most recently, during her time at Google Brain, she explored and quantified unintended biases learnt from large, messy natural language datasets. She has also organized a for-credit intensive Introduction to Deep Learning winter course taught at MIT, and gives related lectures in various departments.