Sam Anzaroot

Sam Anzaroot is an applied AI researcher at Verneek. Sam previously worked as a principal data scientist at Dataminr, where they helped build out the AI/ data science team. At Dataminr, Sam worked primarily on natural language problems on social media, detecting public safety events. Sam has led multiple high-profile teams and has a track record of successful projects. Some of their notable projects include automatic event summarization using text-generation algorithms including seq2seq LSTM and Transformers, location prediction using neural networks and conditional random fields, topic prediction using fasttext, and language identification using structured SVMs. Prior to working at Dataminr, Sam interned at Oracle labs east, where they worked on scaling up undirected graphical model inference methods by taking advantage of nvidia GPUs, using CUDA. Sam created a GPU version of the belief propagation algorithm written in CUDA and optimized this implementation to allow for 200x speedup in inference and 100x speedup in training of CRFs over sequential implementation. Sam got their start as a graduate research assistant - IESL lab at the University of Massachusetts Amherst.

Sam Anzaroot's educational career includes a BS in Computer Science from Queens College, as well as an MS in Computer Science and Information Extraction from the University of Massachusetts Amherst.