• Trained a model using LSTM to support an AI augmented underwriting system, which recognizes and categorizes the information directly from user’s medical history image.

• This system can process over 50,000 images within 30 minutes, which originally takes an underwriter 1 week to finish.

• Developed a program to extract user’s information from highly unstructured and imprecise txt file using java.

• The program was able to identify and recognize the information from 100,000 files with 99.8% accuracy.

• Developed a program to calculate the real-time parameters of the controller using Matlab.

• The program was proven to control the motor speed in real-time with 1.6% error under multiple situations.

• Obtained highest distinction on a research-based thesis.