Omkar

Reddy Gojala
Boston, MA 02120 | [email protected]| +1 (832)863-0206 | Linkedin | Github | Website
EDUCATION
Northeastern University, Boston, MA May 2021
Master of Science in Data Science GPA: 3.7/4.0
Related Courses: Data Management and Processing, Linear Algebra and Probability, Foundations of Artificial Intelligence, Supervised
Machine Learning, Algorithms

Dhirubhai Ambani Institute of Information and Communication Technology, Gandhinagar, India May 2017
Bachelor of Technology in Information and Communication Technology GPA: 7.7/10.0

TECHNICAL KNOWLEDGE
Programming: Python, R, Java
Database: MySQL, PostgreSQL
Tools: RStudio, Jupyter Lab, Google Cloud Platform, Eclipse, Git, LaTeX, JIRA
Libraries: Scikit-Learn, Tensorflow, NLTK, Tidyverse, NumPy, Pandas, Matplotlib, ggplot2

ACADEMIC AND INDEPENDENT PROJECTS
Text-Speech Generation from Images for Visually Impaired using Neural Networks [source code] October 2019- Present

Car Insurance Fraud Detection using Various Classifiers [Kaggle Kernel] August 2019
• Performed exhaustive feature engineering, mitigated class imbalance using SMOTE and one hot encoded categorical
features.
• Classified data using Logistic Regression, KNN, Decision Trees, Random Forests and Linear Discriminant Analysis.
• Evaluated the performance using Cohen’s Kappa, Area Under the Curve metrics and ROC curves.

Action Recognition using LSTMs and SVM in Pytorch [source code] May – June 2019
• Employed transfer learning to extract image features from a pre-trained VGG network with 16 layers and built an LSTM
with these features for predicting the action being done in a video.
• Built a competing SVM model and analyzed performance of both models.

Yelp Restaurant Reviews Aspect based Rating using Latent Dirichlet Allocation [source code] March – April 2019
• Performed sentiment analysis and predicted top positive and top negative reviews for restaurants
• Calculated aspect (food, service, ambiance) based rating for each review using Latent Dirichlet Allocation by employing
guided seed words and unsupervised pruning to significantly reduce the convergence time and improving accuracy
• Computed overall aspect ratings of each restaurant by combining the individual aspect ratings of each review with
probabilities generated by LDA algorithm
• Recommended top 5 dishes for a restaurant using the most frequent dishes mentioned in the reviews

PROFESSIONAL EXPERIENCE
Center for Complex Network Research, Boston, MA June 2019 – Present
Research Assistant
Objective: Our goal is to use openly available power utility outage data across US to answer important questions like estimated
power restoration time during crisis and natural disasters and factors affecting it
• Collecting time-series power outage data across all states in US, storing it in a PostgreSQL database and hosting the data
on an outage map using Flask web framework

Juniper Networks, Bangalore, India June 2017 – December 2018
Software Engineer
• Designed and developed an Automation Test Framework for testing all flagship products of Juniper
• Took ownership to implement and deliver a multiprocessing paradigm in Automation Test Framework resulting in a 40%
decrease in testing time and received “Juniper Spotlight: Best Performer of the quarter” award
• Initiated and developed a template engine to reduce workload and time to write test scripts by testing teams

The Apache Software Foundation May – August 2016
Google Summer of Code Intern
• Designed and developed a Python utility library podaacpy to access data from NASA PODAAC, the library has been
accepted by NASA Open Source repository and currently has 60,000 downloads
• Integrated library into Apache Open Climate Workbench (OCW) to support a new ocean data source