Rajath Shetty

Boston - admygo@r.postjobfree.com - 617-***-**** - linkedin.com/rajathshetty - github.com/rajath95 - rajath.xyz OBJECTIVE

Computer Science graduate with 2 years of experience looking for a role as Software Engineer SKILLS

Languages: Python, Java, MySQL, HTML, CSS, Javascript Skills: Machine Learning, Backend Dev, Scripting, Full Stack, Algorithms Frameworks: React, Django, Tensorflow, REST API,Mongodb, Nodejs, Scikit, Pandas Tools: AWS, Git, Bash

EXPERIENCE

Jeevahealth.ai — Fall Intern SEPT 2020 - DEC 2020

Was Involved in building a chatbot. Increased accuracy of chatbot by 90% by using a BERT Model trained on a communication corpus to predict sentiment of the user. It used pandas, scikit, and huggingface libraries RoundSqr Inc — Backend Developer FEB 2019 - JULY 2019 Contributed towards the development of Django views and REST APIs to develop the backend of a commercial Android Application. Including User Management and SMS Gateway. It made use of REST API, Django framework, MySQL database and hosted on AWS EC2 Instance. It contributed to a revenue of 50,000$ Eautomaton Cognitive Services — Python Developer SEPT 2017 - SEPT 2018 Worked on building an end to end Machine Learning based Real time Application. It was a Predictive Engine deployed in Hospitals. The backend was built using Django and the front end used HTML, CSS and Javascript. It used Machine Learning models based on Random Forest algorithms that trained on demographic details as well as factors like weather, location to make predictions. It contributed to an annual revenue of 210,000$ for my startup. EDUCATION

University Of Massachusetts, Boston — Masters in Computer Science August 2019 - May 2021

Bangalore Institute Of Technology, Bangalore — Bachelor of Engineering June 2013 - July 2017

PROJECTS

Sentiment Classifier

● Classified IMDB comments into positive or negative sentiment using TensorFlow, NLTK and Gensim libraries

● Converted labels into one-hot encoding and used NLTK library to remove stop-words and tokenize words

● Fed tokens to Word2Vec model to convert it into word embedding; used conventional gradient descent for loss function Make a Deal

● Mini web game built with Javascript that asks user to pick the right door

● Winner gets rewarded with a prize