Vraj Shah Email: vraj.shah@rutgers.

edu
Portfolio: vraj152.github.io Mobile: +1-732-640-6384
Github: github.com/vraj152 LinkedIn: vrajshah152
Education
Rutgers University New Brunswick, NJ
•
Master of Science in Computer Science; GPA: 3.9 Jan. 2020 - Dec. 2021
Courses: Design and Analysis Of Algorithms, Machine Learning, Computer Vision, Intro to Artificial Intelligence, Massive Data Mining

L.D. College of Engineering Ahmedabad, India

•
Bachelor of Engineering in Computer Engineering; GPA: 3.54 (8.85/10) July. 2014 – May. 2018
Courses: Data Structures, Analysis Of Algorithms, Software Engineering, Web Technologies, Database Management Systems

Skills Summary
• Languages: Python, Java, C++ (Basic), JavaScript, jQuery, MySQL, HTML
• Frameworks: Flask, Django, Spring Boot, Spring MVC, Hibernate
• Tools: GIT, JIRA, Eclipse IDE, PyCharm, Android Studio, LATEX, ServiceNow, ClarityPPM, Docker, Kubernetes
• Other Skills: AWS, GCP, Web Services, REST API, AJAX, Mathemetics(Probability, Statistics), Web Scrapping
Experience
Google Seattle, WA
•
Software Enginner II February 2022 - Present
◦ Currently working in Google Kubernetes Engine (GKE) team.
Boston Consulting Group - GAMMA Chicago, IL
•
Software Developer Intern September 2021 - December 2021
◦ Working on Retail Catalyst product which helps clients strategize pricing and facilitate simulation runs.
◦ Developed new features like integration with Slack where only error messages will be posted, which makes the debugging
easier for the developers. Enhanced the robustness of the application by resolving the identified issues and minimized
the frequency of error messages.
◦ Also migrated the backend from Flask to FastAPI.
Royal Bank of Canada - Capital Markets New York, NY
•
Quantitative Technology Summer Analyst June 2021 - August 2021
◦ Worked on Credit Model to predict the inter dealer prices for municipal and corporate bonds.
◦ Implemented an ETL pipeline to automate data preparation, non-linear kernel transformations and feature extraction.
◦ Trained Linear Regression, KNN, Random Forest, SVM and MLP models and fine-tuned their hyperparameters using
coarse and fine grid searches.
◦ Accomplished 14% improvement in the performance of a model over existing model.
Rutgers University, Computer Science Department New Brunswick, NJ
•
Graduate Research Assistant January 2021 - June 2021
◦ Data Puzzle Generator: Working on SAS funded project to design & develop E2E Data Puzzle Generator Tool for
Data-101 class at Rutgers University under guidance of professor Tomasz Imielinski. Using Random Forest and Decision
Trees algorithms. Developed and released the application in 5 months, which reduced the puzzle generation time by
70-80%.
Accenture Pune, India
•
Associate Software Engineer July 2018 - November 2019
◦ Client: Credit Suisse (Pune, India)
∗ OnePPM: OnePPM is a project and portfolio management tool which is built on top of Clarity PPM, which enables
you to manage different projects and portfolios running in your organization. Customized this tool based on
requirements given. Developed several modules related to timesheets and expenses filled by employees or contractors.
∗ Amelia: Was part of the research team. Our goal was to reduce human efforts required in solving repeatedly raised
issues by automation. Used NLP to make chatbot more efficient. It helped to save around 20 man-hours per week.
Projects
• Job Tracker: September, 2020
◦ Developed an application where users can keep track of the job applications.
◦ Wrote scripts to web scrap the necessary details from job boards like Jobs.lever, Greenhouse.io, Workday etc.
◦ Designed and developed web application where user can register and add positions where they have applied, update the
status and search through applications. Worked on RESTfull Web Services, used MySQL database.
◦ Hosted application on AWS. Used services like RDS, API Gateway, EC2, SNS, Lambda.

• Shortest Path on Google Maps Using A* Algorithm: May, 2020

◦ Implemented A* algorithm to find the shortest path on Google Maps.
◦ Required Graph Structure was extracted from OpenStreetMaps containing ∼27k nodes and ∼63k edges.
◦ Algorithm finds the path which is suggested by Google when you select travelling mode as Walk. Although graph is
dense (O(bd )), algorithm produces output within 5 seconds.