Tejas Umesh

Work Experience

June 2024 – June 2026

Dignity Health, Arizona, USA

Research Assistant (Data Engineering)

Behavioral studies have consistently shown that whereas we often remember the gist of emotional stimuli, we forget the details. While research on this phenomenon has focused on well-known brain regions like the amygdala and hippocampus, my project analysed brain signals from the pulvinar, a lesser-studied region known to be involved in visual and emotional processing.

April 2022 – May 2024

Decision Theater, Arizona, USA

Data Analyst (AI Modelling)

Developed an LSTM time series forecasting model for predicting Valley Fever from weather data in Maricopa County, AZ. Achieved a low mean squared error (MSE), identifying optimal prediction windows upto 30 days. This project aims to help doctors and researchers predict patient load for each season, which helps prepare better for this disease

Aug 2020 – Aug 2021

Robert Bosch Centre for Data Science & AI, India

Research Assistant (AI Engineering)

Led a team of four to develop a lung tumor detection system using SVM and Random Forest, achieving 83% accuracy. Conducted feature analysis with random forest to identify the top 10% tumor aggression features. This healthcare project aimed to increase the speed and efficiency of tumor detection as a preliminary screening tool.

July 2019 – July 2020

Indian Institute of Science, India

Research Assistant (AI Engineering)

Implemented a BLSTM model for speech-to-EMG (facial muscle signals) mapping, with the accuracy metric between the generated facial signals and the ground truth being 76%. This research aimed to improve previous efforts in visualisation of facial muscle activation from speech signals, and could potentially improve EMG-Speech mapping which could be helpful for people suffering from speech disorders like Laryngectomy.