Hi There!!
Thanks for reaching out to me!!
My Name Is Rakshit Sinha.
A data-driven Storyteller and an aspiring Data Scientist.
I can tell you how you can triumph over your business rivals using in-depth data analytics
MY SKILLS
I have been more than 3 years in Data Science industry, and have helped many clients to grow their business on the basis of extensive data analysis.
About Me
What’s up, folks! Welcome to my portfolio website.
Currently, an MS in Information Systems student at Robert H Smith School of Business at the University of Maryland, College Park.
I have worked as a Data Analyst for more than three years and I’m really passionate about discovering useful business information using various statistical learning methods and then conveying the “Data-Story” using various visualizations and animations.
Achievements:
- Runner-up in “UMD SAC Datathon-2022 organized by Deloitte” at UMD (College Park)
- Runner-up in “Best Cybersecurity AI Project” category in “Info Challenge” at UMD (College Park)
- Best 15 Hacks in Hacklytics-2022 at Georgia Tech
- Terrapin Scholarship, Recipient
3
Years of Experience
2
Awards & Honors
My Projects
Detecting Pneumonia by Chest X-ray images
- Built CNN model to detect whether patient suffering from pneumonia or not by using chest X-ray images.
- Created “filter” using auto-encoder Convolution Neural Network, resulting in recognition of only relevant sections of images in high resolution leading to reduction in image size and memory allocation.
- Decreased process time (for model creation) from three hours to 40 minutes, and increased accuracy from 43.6% to 79.8%.
Predicting “Heart Stroke” probability of patients suffering from COVID-19
- Constructed classification model to predict the probability of suffering from heart stroke for COVID-19 patients.
- Increased balance of dataset by removing high bias towards one target label using “resampling method”.
- Reduced dataset bias from 96% (towards patients not suffering from stroke) to 50%, and increased F-1 score from 19% to 73% by implementing Random Forest algorithm by considering various socio-medical parameters.
Predicting Air Quality Index by using Satellite data and pollution sensor data
- Developed a regression predictive model to estimate the air quality index for the city of New Delhi by statistically analyzing various geographical, industrial and agricultural trend around the region which was accurate up to next 5 minutes.
- Detected sudden and anomalous spike in air quality index which was caused by the stubble burning in nearby regions, and depending on this, Increased the R-squared score from 37% to 59% by doing various feature engineering
Made the model dynamic by deploying it on Web using Django and service platform (Heroku)
My Experience
Positioning myself as an Analyst with HCL Technologies for more than three years, I have leveraged my expertise in analyzing big data using various machine learning algorithms and technical understanding of Python, R, and SQL to not only work on massive datasets but also draw meaningful insights and elucidate them by creating interactive dashboard visualizations using tableau and Power BI
As a Machine Learning Intern at GPS Desk, I learnt how to apply various ML algorithms to develop a highly functional and cost-effective model. Despite the initial challenges of processing large datasets and deciding on the perfect algorithm, I was able to implement the correct predictive model after a lot of trials. With the help of my mentors and teammates, my first professional venture into the world of data science turned out to be extremely fruitful.
Work Experience
May 2022 – Present
Data Science Intern – Optimal Solutions Group, LLC (College Park, MD, USA)
- Developed and Deployed (on AWS) an image classification model for NARA (National Archive and Record Administration) using Tensorflow’s NasNet architecture capable of classifying images into 11 different categories with an overall accuracy of 88.56%.
- Developed an application to increase the ‘user penetration’ of the newsletters published by “Optimal Solutions Group” using advanced NLP techniques (like BERT) to increase the ‘click through rate’ from 7.4% to 25.8%. The NLP model returned the likelihood of a newsletter being classified as ‘spam’ and recommended words to be replaced.
- Built a Web crawler REST API capable of web scraping as it crawled from one link to other using libraries like AutoScraper, urllib, fake_useragent, Flask, bs4 etc. Due to the time delay, the API avoided being detected as a bot and efficiently crawling through multiple pages, ultimately giving results in json format.
May 2019 – August 2021
Analyst – HCL Technologies Pvt. Ltd. (Noida, UP, India)
Job Responsibilities
- Conduct Predictive Analysis to forecast the total memory utilization percentage for the servers present in the client’s environment and explain the inferred data to the client through data visualization.
- Coordinate with Managers to translate business requirements into coherent Business Intelligence (BI) reports.
- Generate BI dashboard and reports with the help of various BI tools and create functional and efficient BI data models for clients
- Mentoring Junior Analysts in BI report and dashboard creation.
- Performed statistical analysis to detect outliers and anomalies.
Awards and Honors
- Inducted into NORAM-DWP Hall of fame for 2020-H1 (Apr to Sep 2020) by the Global Domain Lead.
- Received Certificate of Appreciation from the client for outstanding work and the Excellent Performer Award for FYI 2019-20.
July, 2018 – May, 2019
Graduate Engineer Trainee – HCL Technologies Pvt. Ltd. (Noida, UP, India)
Job Responsibilities
- Analyzed huge sets of data to determine the trend by implementing various ML algorithms for inferring information and carried out Predictive Analysis.
- Presented key insights to the client through data visualization tools, such as Power BI or Tableau.
May, 2017 – August, 2017
Machine Learning Intern – GPS Desk (Bangalore, KA, India)
Job Responsibilities
- Applied Machine Learning algorithms to achieve a highly functional, efficient and cost-effective model for developing a fuel calibration system.
- Used large data of different vehicles and predicted the exact fuel calibration required using data cleaning, featuring engineering and polynomial regression to set up the correct data model and explained it to my manager using data visualization techniques.
Accomplishments
- Received a certificate of appreciation from the senior management
- Gained in-depth knowledge of pragmatic machine learning applications and data pre-processing
Robert H Smith School of Business – University of Maryland, College Park (December, 2022)
MS in Information Systems
- Terrapin Scholarship, Recipient
- GRE: 322 (QA: 169; VA: 153; AWA: 5)
R.V. College of Engineering, Bangalore (May, 2018)
Bachelor of Engineering in Electrical and Electronics Engineering
- CGPA: 7.97/10 (Degree Conferred in March 2019)
- Reached Quarter-Finals in Hackathon ‘HackTrack’- 2016 at IIT-Bombay
- Contributed as a Core Committee member of E-cell and organized Mock World Humanitarian Summit in college and received Excellence Award for exceptional endeavors; 2016
- Organized Quiz Fest in college as a core committee member; 2015
“The world is one big data problem.”
Andrew McAfee
(Co-director of the MIT Initiative)
“Without Big Data Analytics companies are blind and deaf wandering out onto the web like a deer on a freeway”
– Geoffrey Moore
(American Management Consultant and Author)
If you have made it this far, then it seems like you are interested in my portfolio. Want to team up for a Kaggle competition or a Datathon or any data challenge? Mail me at rakshit.sinha@marylandsmith.umd.edu
Are you a recruiter? Drop a Hi if you really like my profile, on My LinkedIn Profile, or Mail me at rakshit.sinha@marylandsmith.umd.edu
Thanks again for your time!! Really appreciate it!!