Karthik Subramanian

Welcome!

I’m a 4th year undergraduate Computer Science student at Georgia Institute of Technology with focus on Artificial Intelligence and Modeling/Simulation. I’m from Belle Mead, NJ. Check out my awesome work!.

As the amount of data being produced per day is increasing rapidly, it’s interesting to be able to derive meaning from data in order to make informed decisions. As a result, I became interested in using data science and software engineering to help solve challenging problems.

Current Experience

1. Undergraduate Researcher in Automated Algorithm Design Team (January 2021 - Present)

• Objective of Automated Algorithm Design is to use genetic programming and ML techniques to create an automated method that develops hybrid algorithms. We do this through an automated machine learning framework called EMADE (specific to GTRI). Genetic Programming takes the best human solutions and tries to evolve hybrid algorithms to outperform existing methods

• During both Spring 2021 and Fall 2021 semesters, I was part of the NLP subteam. This subteam's goal is to use EMADE to produce models to solve specific NLP problems (and outperform the existing algorithms). During the Spring 2021 semester, I worked on sentiment analysis using the Amazon Review Dataset from Kaggle. During the Fall 2021 semester, I'm working on developing a Question-Answering system using the SqUad dataset and EMADE. I've implemented BiDAF architecture into EMADE to help the team make significant progress

• Link to Information on Automated Algorithm Design

2. Software Development Engineer Intern at Amazon (May 2021 - August 2021)


• Developed a Python package to personalize datasets of 100000+ rows to enhance functionality of a healthcare application

• Improved runtime of package by 97.5%

• Advocated for better unit testing practices using parameterized testing

3. Undergraduate Teaching Assistant (January 2021 - Present)

• I'm a Teaching Assistant for a course called Objects and Design. The objective of this course is to introduce the software development life cycle and methodologies in which teams work to develop a software application.

• I explain concepts in Object Oriented Analysis, Object Oriented Design, Iterative Development, Agile methodology, SOLID/GRASP principles to 500+ students along with reviewing course material for consistency with the expected course objectives and grading exams

• During the Fall 2021 semester, I manage and organize an online forum called Piazza, where students can ask questions about the course material. Through the questions students ask on Piazza, I'm able to pinpoint areas of confusion and help to create material and resources to help students understand the material in a simpler way

Past Experience

1. Georgia Tech On-Campus Notifier (December 2020 - January 2021)

• Problem: Most on-campus jobs at Georgia Tech are posted on this website. Students would have to manually look through these job postings for open positions, but vacancies fill quickly and there isn't a way of getting notified as soon as vacancies open.

• Solution: I built an email notification service with a few of my friends. This application sends an email notification to subscribed users as soon as a new vacancy is posted

• How the application works:

  • Scrape the Georgia Tech On-Campus Job Board through Beautiful Soup API
  • Compare newly found job postings to an existing database of job postings on Mongo DB
  • If there is a new posting, then our application sends an email notification to an email list stored on Mailchimp API with the help of Gmail API

• Results: As of Jan 16th 2021, there are 300+ subscribers




2. Farm Simulation Application (August 2020 - December 2020)

• Leveraged Agile Principles and Iterative Development to develop an interactive farm simulation application using Java FX in a team of 5.

• Implemented OOP principles to develop framework, extensively tested codebase using Junit framework

• Check out my video submission

• Results: Nominated for top 10 best projects out of 100+ projects




3. Stock Analysis Project (February 2020)

• Problem: Social media is very pervasive in our daily lives. The price of a company's stock to some extent could be influenced by the public's opinion. I was interested in seeing if social media posts on a given company could be used to predict the opening price of the stock on the next day

• Solution: I developed an algorithm that scrapes Reddit data on a given company along with historical stock prices. Our algorithm performs natural language processing in order to get the data into a format that's usable for a machine learning model.

• How the algorithm works:

  • Scrape Reddit posts on a given company's subreddit through the Pushshift IO API
  • Scrape historical stock prices on the company using the Alpha Vantage API
  • Pre-process the financial data with Pandas
  • Convert the Reddit posts into vectors through the SpaCy toolkit on a dataset of at least 30000+ rows
  • Match up the vectorized Reddit posts and the financial data on the basis of date
  • Tried out models such as decision trees, Random Forests, XGBoost and decided that Random Forest gave the best performance

• Results: Acheived accuracy of 85-95% with mean absolute error of $33.70 on a dataset with standard deviation $69

  Link to Project

About

Computer Science fascinates me due to its wide applicability. I enjoy the power of computing to make meaning out of large amounts of data to drive informed decision making. Currently, I'm involved in a student organization called Data Science Club at Georgia Tech. The main mission is to make data science accessible to those new to the field. I drive this mission forward as Director of the Content Team. I design educational material in the form of interactive Python notebooks for a 10-12 week long virtual bootcamp to illuminate the beauty of data science and present this content to 50+ participants during workshops. Check out our website for more information.

In my spare time, I'm actively exploring new applications of Computer Science through side projects and hackathons. Learning through a project oriented mindset drives me. I also like to play card games and I'm willing to pick up any new card games. I also watch intriguing TED-talks and am a fan of logic puzzles (I'm currently working through "Alice in Puzzleland" by Raymond Smullyan). I'm currently learning to cook and enjoy the process.

I'm currently exploring Medium as an outlet to share my thoughts. For now, I'm writing about topics within probability and statistics (but presented in an easy to understand way). You can check out some of my articles at Medium Articles

Reading

This page contains books that I'm currently reading and my personal thoughts. Feel free to message me about any books that you are reading

I'm currently reading this book. It's a very eye opening piece and really shows how one can excel in a skill and consequently derive passion from that. I personally enjoy the way that Cal Newport explains his thought process on what it takes to achieve excellence and this novel has impacted the way I approach my work. I personally recommend this book to anyone who wants to learn something new.

My Treat 😄

This page contains some miscellaneous information about myself. I hope you all enjoy it!

This video is one I took with my friend, Oliver, during Summer 2021 in Seattle. We wanted to put together this video to commemorate our trip and create a lasting memory.

The above pizza is one that my roommate and I made for a dinner. We used Trader Joe's cauliflower crust pizza and added toppings of olives, mushrooms, and bell pepper

The above photo is Halwa, one of my favorite Indian desserts. I'd like to thank my mom for taking the time to prepare this dish for my 21st birthday.

The above photo is a quesadilla my roommate and I made together. We stir-fried bell peppers with fajita and taco seasonings. Refried beans were seasoned with taco sauce, red chilli powder, coriander-cumin powder and a hint of oregano. Guacamole was there to accompany the dish!

Delicious Thai curry my roommate and I made together. Ingredients used included but not limited to sriracha sauce, tofu, bell pepper, green chilli, onion, turmeric, cumin, ginger garlic paste, carrot, black pepper powder, kitchen king spice mix, garam masala, coriander cumin powder, red chilli powder, salt

Elements

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';