Prakul Resume

Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

PRAKUL

716-507-6412 | [email protected] | https://www.linkedin.com/in/prakul-95/


EDUCATION
Master of Science: Data Science, University at Buffalo, The State University of New York, August 2020-
February 2022(CGPA 3.750)
Bachelor of Engineering: Electronics & communication, Guru Gobind Singh Indraprastha University, India,
August 2013-July 2017 (73/100)

SKILLS & TOOLS


Languages: C, Dotnet C#, Python, R, Visual Basic, bash scripting, command line.
Data Management & Analytics: MSSQL, SQLite
Web Development: HTML, Web development dotnet C#
Tools: Jupyter, Visual Studio, Matlab, FCP

DATA SCIENCE PROJECTS


Predicting Malicious URLs using Lexical analysis: Python, Jupyter
• Cleaned the raw dataset using exploratory data analysis, performed various experiments to tune
parameters and optimize model to predict if the URLs are spam or not with an accuracy of 93%
Map-reduce using HDFS and Python: HDFS, web API, Jupyter, python Tkinter
• Created map reduce programs for co-occurrences, shortest path and page rank programs for both
small and large-scale datasets fetched from web APIs by utilizing Hadoop file system.
Sentiment analysis of clustered tweets from Twitter data: syuzhet, dplyr, Twitter API, R, K-means
clustering, ggplot2, wordcloud
• Fetched, cleaned and munged apple tweet data from Twitter using TWITTER library in R
• Applied K-means clustering to the data and plotted the sentiment analysis graph for each cluster.
Creating a new Machine learning model to replace COMPAS to minimize racial discrimination: Python,
Keras, tensor flow, command line.
• Built and compared 15 machine learning models (3 models (SVM, Neural net, Naïve Bayes) extended
to 5 each) by optimizing- Demographic parity, Equal opportunity, Maximum profit, predictive parity and
single threshold techniques.
• Built a detailed report for choosing the desired model which was cost effective and had threshold
parameters such that it reduces racial discrimination among ethnicities. The report was awarded
highest marks for the detailing.
Prediction of Mental health using Classification Algorithms (in Python): Random forest, LDA, sklearns,
numpy, pandas, seaborn matplot, metrics.
• Cleaned, munged, and applied machine learning algorithms to classify and predict mental health of
participants. Accuracy 88%
WORK EXPERIENCE
Application services specialist, NTT DATA, Bangalore, India: December 2017 – Mar 2019
• Supported 55+ applications by providing L1-L3 backend support. Tasks included data warehousing,
incident management from P1-P3 and change management through ITIL. Certified in ITIL and .net
development.
• Created automated scripts to extract client reports which helped cut down team effort.
• Created root cause analysis report after incident resolution.
Software developer trainee, NTT DATA, Delhi, India: August 2017 –December 2017
• Developed client-side login encryption for employees and managers which helped in re-enforcing the
security protocols of the company. This application was used throughout the company in India.
• Tested the Skill matrix application for functional errors.
• Created a functional CD rental website where users can create a login ID and password and rent, buy
and return CDs. Awarded Best application among the training class.
AWARDS & INVOLVEMENT
Certificate of training, Telecommunication dept., Indian Railways, Jun 2016 – Jul 2016

You might also like