Kishan Thakkar

(847) 975-1783 · kishanthakkar28@gmail.com

I am currently a junior at the University of Illinois Urbana-Champaign studying Computer Science. I am experienced in backend software development , android/web development, and data analysis. Click around through my website to learn more about my experiences, interests, and more!

Experience

Pinterest

Software Engineering Intern
  • Built a data workflow, using a series of Apache Spark jobs, that would identify Pinterest pages crawled by the Googlebot, determine interlinks originating from these pages, and calculate aggregated metrics about the crawled and linked pages.
  • Utilized the data workflow to collect one month’s worth of crawl data . The workflow identified over 70 billion Pinterest pages crawled by the googlebot and over 200 billion interlinks originating from these crawled pages.
  • Reduced the overall run time of the data workflow by 40% by eliminating data skew which allowed the Spark jobs to run with a greater degree of parallelism.
  • Leveraged the Spark GraphX library to build a directed graph representing the Pinterest pages identified from the data workflow and the interlinks between them; Used GraphX to also run PageRank on this graph which assigned a score to each page in the input graph allowing us to rank the pages from ”best” to ”worst”.
  • Used inferential statistics and correlation analysis to identify relationships between the generated ”ranking” of a page and the amount of traffic that page received.
May 2020 - August 2020 Scala, Python, Apache Spark/SparkSQL

Amazon.com

Software Development Engineer Intern
  • Designed and bootstrapped a cloud-based in-memory database which served as a read-replica for a NoSQL database fundamental for AWS Batch job scheduling.
  • Developed an “event applier worker” application which maintained eventually consistency between the two databases by propagating changes from the NoSQL database to the in-memory database.
  • Used Java multi-threading to increase throughput of the worker by 50%; maintained a thread safe application by using appropriate design patterns and data structures.
  • Engineered a monitoring protocol which would trigger the appropriate failure recovery mechanism depending on a complete or partial failure of the in-memory database.
  • Iteratively developed modularized packages while writing JUnit tests and logging using Log4j2.
May 2019 - August 2019 Java, AWS, Redis, NoSQL

Education

University of Illinois Urbana-Champaign

Bachelor of Science
Computer Science

GPA: 3.76

August 2017 - December 2020
Relevant Coursework:
  • Algorithms & Models of Computation
  • Data Structures
  • Systems Programming
  • Deep Learning
  • Artificial Intelligence
  • Distributed Systems
  • Parallel Programming
  • Computer Architecture

Projects

NLP Stock Predictor

Python
  • Performed sentiment analysis on Reddit comments about Fortune 500 companies using TextBlob.
  • Built a regression model to find trends between the ”polarity” (change in sentiment over the course of a week) for acompany and its historical change in price during that time span.
Python, Pandas, Numpy, Scikit

ESPN Fantasy Football Web App

Python
  • Flask web-application designed to import ESPN fantasy football data and provide feedback/analysis for individual teams.
  • Created client that allows authorization from ESPN to access data from private leagues.
  • Improved website performance by incorporating Memcached, a distributed memory object caching system.
  • Used Pandas to organize individual player fantasy data from CSV files into a flexible and efficient data structure.
  • Implemented responsive, interactive features/front-end UI using using Bootstrap.
Python, Flask, Pandas, Bootstrap

Skills

Programming Languages & Technical Tools
  • Python
  • Scala
  • C
  • C++
  • Java
  • Apache Spark/SparkSQL
  • Android
  • NoSQL
  • Amazon Web Services
  • CUDA

Interests

Apart from being a student, I enjoy working on independent programming projects, playing chess, and reading. I am also a huge NBA (Go Bulls!) and NFL (Go Bears!) fan and enjoy looking at trends in player and team data.

Otherwise, I am usually playing chess Challenge me! or reading data driven articles fivethirtyeight.com.