thakkarparth007


I'm Parth, an MS student at UIUC, exploring how to apply Machine Learning and Program Synthesis techniques to make software development easier. If you've heard about Codex or Github Co-Pilot, or FlashFill...or Machine Programming, I'm into that sort of stuff.

Prior to this, I was a Research Fellow at Microsoft Research, India. I had a lot of fun working with Dr. Venkat Padmanabhan on a couple of networking and data science projects. Even before that, I worked with Dr. Senthil Nathan and Dr. Praveen Jayachandran on improving performance and scaling of Blockchains at IBM Research, India.

I am a bit of a speed-typing fanatic, and you can check my profile on typeracer.

I also happen to love making what people call bad jokes. Here's one:

A miser political worker met with an accident, and ended up destroying his cycle. He went to a party co-worker to get it repaired. Whom did he go to?

- The Spokes-man.

Publications

  1. Optimizing Network Provisioning through Cooperation accepted at NSDI 2022.
    Authors: Harsha Sharma, Parth Thakkar, Sagar Bharadwaj, Ranjita Bhagwan, Venkata N. Padmanabhan, Yogesh Bansal, Vijay Kumar and Kathleen Voelbel.

  2. AutoSens: Inferring Latency Sensitivity of User Activity through Natural Experiments accepted at IMC 2021.
    Authors: Parth Thakkar, Rohan Saxena, Venkata N. Padmanabhan

  3. Scaling Hyperledger Fabric Using Pipelined Execution and Sparse Peers accepted at SoCC 2021.
    Authors: Parth Thakkar, Senthilnathan Natarajan

  4. Performance Benchmarking and Optimization of Hyperledger Fabric accepted IEEE Mascots '18 (Best Paper Award).
    Authors: Parth Thakkar, Senthilnathan Natarajan, Balaji Viswanathan

Writings

I have written a couple of articles:

Projects

I've listed some of the important (lol) projects I worked on during undergrad:

  • Generic Applications Portal

    This is essentially an HTML-based DSL along with a lot of extra magic, that makes writing complex form-based applications very simple. The developer writes one file in an HTML-like language, and can specify all the fields, along with their validation rules. He can use a bunch of fairly useful built-in controls, or he can also mix in regular HTML elements in the markup. This is then compiled to generate frontend and backend-related code.

    Using this, writing form-based websites that have 100s of fields becomes very, very simple - as I have experienced while developing the the portal for PhD/MS admissions for our college.

    While the features are more or less complete, proper usage documentation is yet to be added.

  • Dalal Street

    Dalal Street is an online virtual stock exchange event in Pragyan, the techfest of NIT Trichy. Players compete against other players and bots by trading stocks in real time to get rich. The 2018 edition of Dalal Street received participation of 500+ people.

    The core components of the system were the react-based website, android app, bots written in python, GoLang based server (including matching engine), all of which communicated over gRPC.

  • RemBook

    RemBook is like the yearbook for graduating students of NIT Trichy. It is used by both UG and PG students alike. It is one of the most heavy traffic applications in our campus (the image shows about 20% of total traffic). This is a single-page application, and taught me the importance of caching in such a heavily used system. One of the most satisfying projects I've worked on!

  • Rest of the projects can be viewed on my Github.

Contact

You can email me on:
thakkarparth007 at gmail dot com

I'm fairly active on Twitter.

My LinkedIn profile can be found here.

My Github profile is here.

Resume

My resume can be viewed here.

Copyright © Parth Thakkar