Manasi is a PhD student in the Database Group at MIT CSAIL advised by Sam Madden. She works on interactive systems to enable rapid data analysis using visualization and machine learning. She holds Bachelors degrees in Computer Science and Mathematics from Worcester Polytechnic Institute, and a Masters degree from MIT.

Research Summary

Analyzing data to make decisions, to build new products or to make processes more efficient has become a top priority across industries. However, the growth in data and data analysis use cases has far outstripped the number of trained data scientists and tools for analysis. Manasi’s research involves developing new types of analytical tools to accelerate the process of going from data to actionable insights. She builds systems tailored to the real-world practice of data analysis using a combination of systems optimizations, statistical techniques, and human-in-the-loop design. Towards this goal, Manasi has worked on SeeDB, a novel type of recommendation system for visual data analysis. Given a dataset and analytical task, SeeDB searches through thousands of visualizations to automatically recommend those showing trends of interest. In addition to data visualization, Manasi works on systems to enable fast, iterative machine learning. She is currently working on a system called Sherlock that is designed to support the entire modeling process including efficient building and testing of hundreds of models, management of workflows, and meta-analyses across models.