Yashvardhan Rathi

Working at Weber Shandwick
Studying at University of Texas at Dallas
Living in United States
Website https://github.com/ZlatGod

About

I am currently pursuing a master’s in Information technology and Management at the University of Texas at Dallas with a focus on Business Analytics and Intelligence. I will graduate in May 2020.
A flexible, proactive and curious team-player interested in learning new technologies and services.
I am fascinated by Data Engineering and Cloud Computing fundaments and have industry experience in the same.
I am a certified Mountaineer and a FIDE rated Chess player. Applying my analytical skills in Chess and discipline in Mountaineering to my data-driven career helps me gain a lot of insights into the field.

Experience
• Data Pipelines: Developed a process to ingest and clean the past 18 months (3.5 million rows, 18 cols) social data relating to our client using NLP techniques in Python. Created customized requirements based on changing client needs
• Scripting: Performed ngram analysis, topic analysis, and NER to feature the most intriguing insights to the client to answer detailed questions like What is being said? What does the customer want? What can we change?
• Cloud Engineering: Hosted and automated the data pipeline on GCP to deliver timely reports and insights to our client
• Dashboard: Used IBM Tone Analyzer to analyze tones of texts and create dashboards in Tableau to present important insights.
• Designed production-level data pipelines to enable transfer data-transfer from an OLTP database to a data warehouse
• Implemented ETL (Extract, Transform, Load) process that includes data management via MSSQL in Azure, and MS SQL Server Management, use of Git for source control, Pandas to clean data, Docker to spin up MSSQL containers.
• Developed a versioning system using Python and SQL to organize the client data and present requirements in a methodical manner which resulted in a 43% increase in company efficiency.
Projects
Battle of Attack Detection Algorithms (BATADAL)
Project description
• Compared the performance of various classification algorithms to detect cyber-attacks in water distribution systems
• Tested logistic regression, support vector machines, decision and boosted trees, random forests, k-means clustering, and neural nets algorithms in Scikit-Learn for a different level of speed and accuracy
• Visualized the results using Matplotlib, plotly, and Seaborn to present further scope of the project

Forecasting the pay offered/expected to a candidate based on various factors
• Analyzed data of undergraduates (26416 records and 34 variables) to identify outliers.
• Cleaned data and handled missing values in R using Random Forest (RF) algorithms and MS Excel
• Predicted the outcome variable with an accuracy of 92.13% using Regression trees and performed Data Visualizations using ggplot2 package in R.