Abhishek Kumar

Hi, I am Abhishek, currently a junior year at the University of California, Berkeley majoring in Computer Science and Applied Math. I have various interests including machine learning, exploring mathematics, and cryptocurrency. Some things I love are cold brew, sitcoms, music, reading, watching stocks, and more.


Some of my Projects

Bear Maps

Made a navigation system in Java for the city of Berkeley that could be accessed through a web browser. Parts of this project included implementing A* to find the shortest path between points, returning an autocomplete list of locations, and rastering images of maps to allow for different size web pages as well. The project was for CS61B at U.C. Berkeley

Build Your Own World

Created a video game from scratch that included features such as keyboard input, random map generation, save and load feature. Generating a world and, subsequently, a playable game showed the importance of managing complexity in code and was originally made for CS61B at U.C. Berkeley.

Scheme Interpreter

Built an interpreter that converted parsed code in Scheme into a Python program and executed this representation, returning what the original scheme code would have. This project made heavy use of tree recursion to both parse and evaluation the scheme expressions and was originally made for CS61A at U.C. Berkeley.

Kinect Vision Processing

Using the OpenCV library, I programmed a vision processing system for a Microsoft Kinect onto a Raspberry Pi, where the video stream would first filter objects according to whether they fell within a certain range of colors. After the color filtering had occured, it would detect contours and find the largest object of a certain color. This code was used for as part of a robotics organization where the robot would then move so the object would be directly in front of the robot.

Personal Website

I had made my portfolio site from the ground up, using HTML, CSS, and Javascript. It includes features for screen responsiveness, allowing for a side navigation bar when the screen width is below a certain resolution, and take advantage of jQuery for scrolling to certain parts of the page. I had included some basic nodejs into my website, allowing for modules that provided additional effects wihtout the neccesity of writing substantially more code.

AI Pong Bot

Using the Pytorch library, I had written a machine learning algorithm to play the game pong, using the gym enviroment for the game itself. It had been trained using reinforcement learning, where the bot had continously played itself for hours. In this time it had learned the game of pong and found that following the ball would allow for the agent to maximize its score.



U.C. Berkeley

Graduation Date: May 2022 (Expected)
Degree: Bachelor of Arts in Computer Science and Applied Math
Relevant Coursework: Artificial Intelligence, Data Structures, Differential Equations, Discrete Math, Foundation of Data Science, Multivariable Calculus, Microeconomic Theory, Numerical Analysis, Linear Algebra, Probability Theory

The Wheatley School

Graduation Date: June 2018
Degree: New York State Regents Diploma with Honors



CyberCube Analytics

Role: Data Science Intern
Duration: June 2020 - August 2020
Contribution: Developed a natural language processing platform for identifying potential security breaches among various companies by analyzing and categorizing past breaches as a means to optimize cybersecurity infrastructure.

NASA Ames Research Center

Role: Undergraduate Researcher
Duration: January 2020 - May 2020
Contribution: Studying the behavior of pilots in challenging simulations in order to predict how they would act in real life circumstances by taking into account variables such over shoulder video, viewing angles, altitude, yaw, etc.

Orics Industries Inc.

Role: Engineering Intern
Duration: June 2018 - August 2018
Contribution: Created circuit designs to control motors, pistons, and sensors for factory machinery via ladder logic and then used the designs to construct and program electrical panels.

Brookhaven National Labs

Role: Research Intern
Duration: June 2017 - August 2017
Contribution: Researched under Dr. Abid Malik of the Computational Science Initiative at Brookhaven National Labs for seven weeks and published my findings at the IEEE 2017 Supercomputing Conference.



Capital Analysts at Berkeley

Role: Analyst
Duration: January 2020 - Present
Contribution: Determined a trade signal for the VIX market via the usage of other technical indicators, such as the RSI, ADX, etc. Programmed and backtested an algorithmic trader using the correlations found with the indicators on Quantopian.

Computer Science Mentor

Role: Junior Mentor
Duration: January 2020 - Present
Contribution: Teaching students about data structures and algorithms on a weekly basis by lecturing on relevant topics, completing worksheets with them, and answering any questions they may have in the class.

UC Berkeley College of Engineering

Role: Academic Intern
Duration: June 2019 - August 2019
Contribution: Assisted course staff on a biweekly basis in teaching students about data structures and various algorithms.

VexU at Berkeley

Role: Head of Operations
Duration: December 2018 - Present
Contribution: Using the OpenCV library, I constructed a vision processing system for a Kinect that detected game obstacles through edge detection and color filtering, directed the robot towards them, as well as target flags and shoot at them.

Seed Base Consulting

Role: External Vice President
Duration: September 2018 - Present
Contribution: Lead external relations, alumni outreach, client acquisition, and corporate partnerships for the organization. Led a team on consulting Samsung about attracting developers to their platform and optimizing user experience through extensive market research, analyzing various metrics, and creating an in-depth, mutli-stage timeline.