This is a standing invitation: if you want to talk software¹, I want to talk to you².

• 2026: organized a 308-page journal called SIGHORSE🦓 and printed 66 copies.
• 2025: wrote a famous 6000 word guide to Optimize Purdue CS. Re-implemented video compression and streaming (and JPEG) for a technical exhibit.

• 2026: saying microseconds multiple times a day at work.
• 2025: worked at Neuralink on the god web app that kept the company ticking. Worked at Modular on Mojo to making GPU programming easier.
• 2024: worked on icu4x to implement a new standard. Did programming languages research³.
• 2021, little bit 2023: made a bioinformatics QC tool.

among other things. Read on for more.

• I graduated and am working full-time in San Jose, CA.

• Awarded 'Outstanding Service to the Department by a Student' from the Purdue CS department, decided by student voting, for Optimizing Purdue CS. Photo here.
• Delivered a talk (video here) at BENTO on spoons.
• (Relevant) classes:
  • (Graduate) Operating Systems. Assignments were hacking on Xinu.
  • (Graduate) Principles of Computer Architecture. Topics include so. much. pipelining. and. so. much. branch prediction. Also branch prediction competition, yay!
  • Honors Research on constrained decoding for LLMs. Slides here.
  • Signals and Systems for Mathematicians.
• I finished running the humorous SIGHORSE🦓 journal for Purdue Hackers. Essentially SIGBOVIK for Purdue, if you're familiar.
• Was an organizer for Purdue Hackers; I ran fortnightly workshops on topics such as GPU Programming (~80 attendees!), algorithmic art, and more.
• _{^{Created a barbie-themed programming environment; won 2nd place in the InnovateHer hackathon.}}

• I interned at Neuralink.

• I interned at Modular working on the Mojo standard library, with Chris Lattner of LLVM fame. Nearly all of my work was open-source in the Modular monorepo. I will not be at ETH Zurich for those wondering.
• I attended YCombinator's AI Startup School in San Francisco, California. I was invited to and attended a dinner with a YC partner as well.

• Created Bad Apple in a C Quine for the Catapult hackathon at Purdue; won first place in the Most Creative category.
• I taught an hour-long spontaneous session on theorem proving in Coq at Purdue Hackers.
• I wrote a guide to Optimize Purdue CS. It is ~6000 words. At least two academic advisors plan to share it with their students in upcoming years!
• I taught the Interpreters workshop for Purdue Hackers.
• (Relevant) classes:

  • Graduate Programming Languages. The first half of the course was based on books from the Software Foundations series, while the second half covered Dafny, plus a brief introduction to separation logic. Assignments included:

    • Proving the correctness of a compiler against a reference interpreter for an imperative language in Coq / Rocq.
    • Proving type soundness of the simply-typed lambda calculus in Coq / Rocq.
    • Formalizing an interpreter for an imperative language in Dafny.

  • Graduate Theoretical Computer Science Toolkit. Topics included convex analysis and optimization, spectral methods, concentration inequalities, and discrete Fourier analysis.

  • Graduate Hardware Security. Projects include:

    • Hacking the ZBT WE526 (WE1626) router.
    • Writing a Trusted Execution Environment for the FemtoRV Risc-V CPU in Verilog.
    • Using the Chipwhisperer Nano to crack AES-128 through power analysis.

  • Graduate Software Engineering for Robots. Assignments included working with ROS2 and MoveIt. I presented the HAMSTER 🐹: Hierarchical Action Models for Open-World Robot Manipulation paper. Slides here.

  • Compilers with Tiark Rompf (of Scala fame). Assignments include lots of compilers written in Scala for a Scala-like language to x86.

• I created keymashed, an interactive exhibit at Purdue Hacker's BURST. It contains interesting things, like:
  • a JPEG-like lossy codec written from scratch.
  • a realtime video streaming protocol and associated netcode.
• filed a bug report for the Linux kernel's memory management subsystem.
• employed as a Research Assistant working with Prof. Suresh Jagannathan working on optimizing MRDTs.
• employed as a Teaching Assistant for Data Structures.
• employed as a developer teaching assistant for Programming in C. I created test modules for course assignments.
• (Relevant) classes:
  • Programming Languages. Assignments included:
    • writing lots of interpreters in OCaml.
    • writing an interpreter for a functional language with type inference and checking.
    • verifying simple programs in Dafny.
  • Systems Programming. Assigments included:
    • a malloc implementation. I made mine robust enough to run Chrome.
    • writing a shell with pipes, quoting, subshells, stream redirection (in C).
    • HTTP/1.1 web server with authorization, cgi-bin modules and dynamic module loading (in C).
    • (team project) a graphical system monitor (à la top). (in C++ and ImGui).
  • Analysis of Algorithms. Assignments included:
    • proving the correctness of dynamic programming and greedy algorithms.
    • creating information compression schemes and comparing against entropy.
    • proving the NP-completeness of given problems via reduction to a known NP-complete problems.
  • Independent Study in Embedded Systems with Prof. Douglas Comer, of TCP/IP fame.
    • designed audio streaming system with Raspberry Pis for concurrent audio playback.
    • wrote an RTP-like protocol for streaming audio over UDP.
• Went to the Midwest Programming Languages Summit at UChicago and met many interesting people!
• Attended every PurPL (Purdue Programming Languages) weekly seminar.
• Received the William C. Nylin, Jr. Endowed Scholarship from the Purdue Department of Computer Science for academic merit.

• Google Summer of Code contributor for icu4x. Mentored by Younies Mehmoud from Google; almost all my coworkers ended up being Google employees. Final report here.
• Summer Undergraduate Research Fellow at Purdue University studying Mergeable Replicated Datatypes using Irmin and OCaml.

• Gave an introductory 60-slideRust talk at Purdue Hackers. About 20 people attended.
• I wrote a clean inter-process communication program using POSIX shared memory.
• I participated in the Open Source Contributor Initiative by OpenSearch and worked on the Performance Analyzer component in OpenSearch using Java.
• (Relevant) classes:
  • Competitive Programming 2: topics included segment trees, hashing, tries, graph algorithms, and combinatorics.
  • Computer Architecture: topics included assembly code, digital logic, processor architecture and so much caching.
    • I wrote a just-in-time compiler in Rust for ARMv8 assembly as part of an honors project. It jitted on function call granularity and cached the generated assembly for future invocations.
  • Data Structures: topics included Big-O notation, heaps, binary trees, linked lists, tries, sorts, hashing, graphs, etc.
  • Linear Algebra
• Attended every PurPL (Purdue Programming Languages) weekly seminar.

• (Relevant) classes:
  • Programming in C
  • Foundations of Computer Science / Discrete Math
  • Honors Multivariable Calculus
  • Independent study in the C++11 memory model with Prof. Suresh Jagannathan. I explored tactics for designing performant and correct lock-free algorithms. Report available on request (until I get around to publishing it).
• Attended nearly every PurPL (Purdue Programming Languages) weekly seminar.

• I was a Rise World Finalist.
• I started working on Librarian with the Babraham Bioinformatics Institute in 2020. My work on Librarian was published here first in 2022, and then the revision in 2024.
• I briefly worked with Trestle Labs on computer vision using OpenCV in winter 2023.
• I was a competitive programmer and qualified for the Indian National Olympiad of Informatics (INOI) in 2020, 2021 and 2022. I studied Antti Laaksonen's Competitive Programmer's Handbook.