Keita Kurita

Keita Kurita has held various roles in the tech industry since 2016. In 2016, they were a Data Scientist Intern at So-net Media Networks Corp, where they developed a geo-targeting system for targeted advertisements and created a parallelized crawler in Python and visualization tool using D3.js for a user browsing history text analysis pipeline. In 2017, they were a Software Engineer Intern at Microsoft, where they analyzed user input data and developed a novel algorithm for error-tolerant user text input prediction, reducing memory usage by 80% and improving speed by over 900%. Keita also developed a software system for experimenting with a neural network based Japanese IME in Tensorflow. In 2018, they were a Data Scientist at So-net Media Networks Corp. In 2019, they were a Software Engineer Intern at Salesforce, where they worked on the Ofek Data Science team on improving Einstein Behavior Scoring. Keita developed an approach to using Learning to Rank (LTR) as the core algorithm and integrated an LTR model into the production pipeline leveraging Mongo DB, Docker, and XGBoost, improving NDCG by upwards of 25%. Keita also extracted features from emails to improve AUC of the core model by upwards of 50%. Currently, they are a Machine Learning Engineer at Robust Intelligence.

Keita Kurita completed a Bachelor's degree in Electrical Engineering and Information Communication from the University of Tokyo in 2018. Keita then went on to pursue a Masters in Computational Data Science from Carnegie Mellon University's School of Computer Science, which they completed in 2019.