develop a generalized language model that could be used for univariate and multivariate time series data across various possible domains that are not just language data, specifically focusing on financial and astronomical time series data

propose a possible pipeline of confidence measures to improve the usefulness of the language model by deriving confidence interval using Bayesian inference

Implemented load-balancing strategy for Azure storage backup users, improving storage limit from 550 TB to 200k TB

Designed storage sharding product design document in a team of 6 people, and collaborated across two teams

Skills used: Go, SQL, MySQL database, Microsoft Azure, grpc

Collected a video Q&A dataset with 400 social content and developed related research questions

Developed a multi-agent computational cognitive model of theory of mind reasoning as the primary data researcher and team leader; used PySWIP to write the production system built on the model

Designed an education application that provides visual scaffolds for students to improve their theory of mind reasoning skills using gamification and user-centered software design in a team of 5 people; successfully tested on 150 middle school students