Rahul thaker

About Me.

I have completed my Master's in Information Management at University of Illinois at Urbana-Champaign. I have completed my BTech in computer engineering and I love to produce results through technology to bring about useful changes, my work has been in statistical modelling, analytics, Machine learning and deep learning.
During my course in UIUC, I implemented state of the art deep learning models on HAL clusters with National Center for Supercomputing applications , and I am currently working with Caterpillar Inc to produce insights and machine learning results for over 10 million customers globally , I have worked with different aspects of the data science umbrella. Apart from that, I also love music and like to play the piano

Abilities

Object Oriented Programming

I have learned and worked with Object oriented programming languages like C,C++,Python etc.

Statistical Analytics

Exploratory and Explanatory analysis, SQL, Decision matrix,t-tests, Hypothesis testing, probability etc.

Cloud technology

Using AWS, Azure, Snowflake and other tools to leverage the power of cloud computing

ML and AI

Machine learning and Artificial intelligence libraries like sklearn, Pytorch, tensorflow. Understanding and usage of Neural networks.

Education

University of Illinois at Urbana-Champaign

Champaign, Illlinois, United States

Masters in Information Management GPA - 4.0

K.J. Somaiya Institute Of Engineering and Information Technology

Sion, Mumbai, Maharashtra

BTech In Computer Engineering GPA - 9.3

T.P.Bhatia College of Science

Kandivali, Mumbai, Maharashtra

HSC Science (2018) Percentage - 86.5%

Jayaben khot high school

Borivali, Mumbai, Maharashtra

SSC (2016) Percentage - 92.8%

Experience

08/2023 - present

Caterpillar Inc.

Data Scientist

Engineered a comprehensive scoring system integrating decision tree models to accurately predict and recommend the most suitable sales channels and product recommendations for over 1 million customers
Performed statistical analysis, generated/collected 200+ behavior-based attributes and developed K-Means clustering models for behavior-based customer segmentation for over 10 million customers globally
Delivered actionable insights for key digital applications and integrated data into Power BI dashboards for real-time visibility into customer behaviors and trends, aiding strategic decision-making using 5 KPIs

08/2022 - 12/2022

National Center for Supercomputing Applications

AI Engineer

Deployed an AI-based teaching assistant chatbot via a multimodal question-answering dialogue system leveraging large language models such as GPT-3, OPT 175B and FLAN-T5 in HuggingFace
Fine-tuned models on custom dataset generated using prompt engineering and data generation through GPT-3
Performed filtering and ranking of response objects by implementing GPT classification models on NCSA servers

08/2021 - 05/2022

Teach for India

Data Engineer (intern)

Streamlined data collection process by directly connecting the Salesforce database to on-prem analytics solution eliminating the need for Excel sheets, saving 20 hours of equivalent weekly manual effort
Automated the process of analysis, visualization and developing reports
Derived new insights and statistical correlations, improving accuracy of selection model by 30%

08/2020 - 12/2020

Pingzee - ZeeQ

Data Science intern

Developed a real-time reporting system for recommending changes based on behavior of infrastructure by leveraging analytics libraries Pandas, NumPy and visuals utilizing scripting layer of Matplotlib
Executed Time-series data analytics and forecasting of backups using SARIMA model with 95% accuracy
Instituted the process of creating highlights, alerts and notifying clients by incorporating SMTP

Extracurricular

Grader for Data Curation course

Grader for CS 598 - Foundations of Data curation offered by the computer science department at UIUC, assisting the professors and grading student answers

Comic-Con 2019 Mumbai

Handled the cosplay management of Comic-Con in December 2019 as a part of the team. I was responsible for handling entries, backstage preparation and stage presence.

CEO of American Chemical Society-KJSIEIT(present)

As the Chief Executive officer of the ACS chapter of my college. I am responsible for organizing inter college events, Interviewing and recruiting suitable members, Communication with members of the committee and college faculty as well as execution of the events.

Organizing Head of CSI-KJSIEIT

As the Organizing Head of The Computer society of India chapter of our college, I was responsible for Scheduling,Guest handling and event planning for different events organized during the Technical festival of our college

Publicity Head of Students council

As the Publicity head of theStudents council I was responsible for the Publicity of different events within the college and throughout different colleges in Mumbai for the yearly Cultural Fest organized within the college.

Projects
Certifications

	  					
	  							NeutraLit
	  						more_vert
	  							Data Visualization
	  							Analysis
	  							Django
	  							Heroku
	  						
NeutraLitcloseA live analytics web-app that gathers data from Twitter and analyzes tweets to provide unique insights. Twitter generates a lot of data through tweets and It would take a lot of time to go through all the trending topics and opinions to stay updated with the current situation. This project does that for you!.

	  								Style Transfer
	  						
	  						more_vert
	  							Deep Learning
	  							GAN
	  							CNN
	  						
				      	Style Transfer
				      	close
				      A model made using GAN and style Transfer to convert real world images to animated style images.  A generative adversarial network is a class of machine learning frameworks Given a training set, this technique learns to generate new data with the same statistics as the training set. 

		  						Airline Sentiment	  							
	  						
	  						more_vert
	  					
	  							Streamlit
			  					Data Visualization
		  						Analysis
	  							Heroku
	  					
				      	Airline Sentiment
				      	close
				      Analysis of Customer reviews to generate insights based on the user inputs. Feedback from customers is an important part for improving any product or service. This web app takes the customer reviews and gives the user an exploratory approach to understanding the reviews and feedback.

Publications

IEEE- Research Paper

Statistical analysis,ETL automation and Intelligence Reporting

June 2022

The integration of data visualization and data analytics has been rising rapidly. Companies working on a large scale have a tremendous amount of data, this data has huge potential and can be used in order to improve the functioning of such an organization. The data can be used to optimize various areas of the organization, improvement of products, giving a better customer experience, increasing efficiency and much more. There has been an increase in the amount of tools and solutions created for working with data. The end result is produced by the combination of various technologies

Cities of India

March 2020

As there is a higher probability for businesses to grow and open a branch in another major city or for employers to have more workforce In another location. Employees not being able to habituate themselves in a certain location or the growth of a business could come to a stop because of a location which may not be suitable. Knowing similarities between locations and analysis of major cities becomes important at this stage.

Metric Comparison

July 2020

Metrics are for Measurement of Models. Different models (regression or classification) require different metrics for example. MSE,RMSE,R-squared,MAE,(R)MSLE etc are the preferred metrics for Regression models. Similarly Accuracy,LogLoss,AUC,Cohen’ s kappa are the preferred Metrics for Classification models. This post is going to give information on How the different metrics stack against each other and when or how they should be used.

Transaction analysis

May 2020

: Banks can have a lot of data generated one of which is transactions. Transaction data for each customer can be used to get a lot of insights about the bank and the customers .Which is what I have tried to do.After doing some exploratory analysis to get some insights .I have done predictive analysis.