Hi, I'm Krishay Garg

Building the Future

Explore Projects Get In Touch
UCLA
STUDENT DEV ID
HOLDER
Krishay Garg
MAJOR & TRACK
Computer Science
CURRENT ROLE
SWE Intern @ Fiserv (Clover)
GPA
4.00
CLASS
2028

Core Competencies

UCLA Logo
Languages
Python, C++, Java, Kotlin, TS, JS, C#
ML / AI
PyTorch, LangChain, TensorFlow, OpenCV, NLP
Cloud & Tools
Docker, Kubernetes, AWS, GCP, Redis, Git
Krishay Garg
AUTHORIZED SIGNATURE

Education & Background

University of California, Los Angeles

B.S. Computer Science | GPA: 4.00 (Computer Science Honors)

Class of 2028

Activities & Leadership

Association for Computing Machinery (ACM) AI Branch Officer
Upsilon Pi Epsilon (UPE) CS Honor Society Member
UCLA Unmanned Aerial Systems (UAS) Computer Vision Team

Work Experience

June 2026 - Present

Software Engineering Intern

Fiserv (Clover) • Fortune 500 Fintech

Engineered distributed systems migrating merchant infrastructure and building event-driven APIs:

  • Engineered a distributed system migrating 500K+ merchants from legacy infrastructure to modern platform. Built Java microservices with automated schedulers to stream records into Kafka pub/sub queue for asynchronous processing.
  • Developed an event driven consumer and Spring Boot REST API to upsert merchants into MongoDB database and trigger activation; deployed Docker containerized services to Kubernetes via Jenkins CI/CD. Built Android features in Kotlin.
Java Spring Boot Kafka MongoDB Docker Kubernetes Jenkins Kotlin Android
Nov 2025 - June 2026

Software Engineering Intern

Stealth AI Startup

Building in stealth...

Aug 2022 - Sept 2025

Software Engineering Intern

League of Women Voters

Led development of community civics technology and automated financial auditing reporting:

  • Built VerifyIt, a full-stack, multiplayer civics education web app using React, Node.js, and Python Flask.
  • Integrated WebSockets to support real-time synchronous multiplayer gameplay and connected to a PostgreSQL database.
  • Scaled application to support 100,000+ active users and presented the platform to millions on NPR (Here & Now).
  • Automated campaign finance report generation from complex Excel files by developing data-processing pipelines in Pandas, NumPy, and Matplotlib, saving staff hundreds of manual hours.
Play VerifyIt Repository NPR Broadcast Finance Report Example
React Node.js Flask WebSockets PostgreSQL Pandas NumPy Matplotlib
Mar 2024 - June 2025

Software Engineering Intern

StemChef

Pioneered augmented reality systems for immersive education:

  • Led the end-to-end development of interactive augmented reality (AR) educational applications using Unity, C#, and TypeScript.
  • Collaborated directly with the CEO to iterate on designs and align engineering efforts with product vision.
  • Mentored and guided incoming engineering interns to streamline code reviews and accelerate features delivery.
Unity C# TypeScript AR Development

Featured Projects

AI & ML

Video Translation Pipeline

An automated pipeline to translate videos into target languages while maintaining precise lip-syncing and speaker voice identity.

  • Applied OpenAI Whisper for ASR and created a custom syllable-matching duration-alignment algorithm utilizing NLP synonym swapping and beam search.
  • Implemented dynamic lip-sync using audio-conditioned U-Nets to manipulate the latent space of pre-trained VAEs.
  • Integrated zero-shot voice cloning using acoustic encoders to preserve vocal profiles.
Whisper U-Net VAE PyTorch NLP
Code
AI & ML

Chess-GPT

Fine-tuned a Large Language Model to output logical and legal chess movements using reinforcement learning.

  • Utilized Group Relative Policy Optimization (GRPO) and LoRA (Low-Rank Adaptation) for resource-efficient alignment.
  • Increased correct move generation by 95%, elevating chess playing capabilities.
  • Achieved a 1000+ rating on simulated games while reasoning through strategic positioning.
LLMs GRPO LoRA Reinforcement Learning Python
Code
LA Hacks Winner

A-Eye: AI Glasses for Visually Impaired

Wearable computer vision powered glasses aiding visually impaired individuals to navigate physical spaces safely. Won 1st Place Track (Voice AI & Synthesis sponsored by ElevenLabs) at LA Hacks 2026.

  • Engineered object detection using YOLO and OpenCV for obstacles and spatial landmarks mapping.
  • Implemented audio translation querying ElevenLabs neural speech processing and OpenAI Whisper.
  • Built mobile streaming client using React Native and TypeScript to forward high-FPS camera streams.
YOLO OpenCV ElevenLabs React Native Whisper
Code Devpost
Web App

VoteClear

An adaptive recommendation platform matching voters with political candidates based on shared values and alignment.

  • Integrated Retrieval-Augmented Generation (RAG) using LangChain to pull structured data on candidates' public positions.
  • Generates custom questionnaire dynamically using LLMs to check alignment on policy nuances.
  • Created lightweight and fast Python Flask backend and styled the UI using HTML5, CSS3, and JavaScript.
Flask LangChain RAG JavaScript CSS
Live App
AI & ML

Diffusion Image Generator

Built and trained a latent diffusion model from scratch, including leading workshops on diffusion concepts for fellow students.

  • Implemented custom denoising U-Net structures and Variational Autoencoders (VAEs) in PyTorch.
  • Supported text conditioning using pre-trained CLIP text encoders.
  • Optimized convergence with parameter-efficient fine-tuning techniques (LoRA).
PyTorch U-Net VAEs CLIP LoRA
Educational
Quiz Bowl App

PowerQB

A text-to-speech quiz practice application used nationwide by high school and college Quiz Bowl teams.

  • Built a robust question sourcing parser pulling from an SQLite database of 50,000+ historical questions.
  • Integrated Google Text-to-Speech API to dynamically read questions with realistic inflections.
  • Acquired 2,200+ active users practicing on academic topics.
Python SQLite Google TTS HTML/CSS
Active

Research & Publications

UCSB SRA

Bayesian Neural Network Topology Uncertainty Analysis

UC Santa Barbara Summer Research Academies (2024)

Researched network topology’s impact on the accuracy and uncertainty estimates of Bayesian Neural Networks. Developed models in PyTorch for image classification on the CIFAR-10 and Fashion-MNIST datasets.

Authored paper submitted to National High School Journal of Science Presented research findings to 120+ seminar attendees
UC Berkeley

Autonomous Vehicle Lane Detection & Navigation

UC Berkeley Autonomous Racing (ROAR) (2023)

Designed lane detection algorithms using computer vision and PID control systems, improving autonomous vehicle performance in CARLA simulation. Placed third in AI Racing competition, published research article on Berkeley website.

Berkeley Article Demo Video Code

Technical Skills

Languages

Python C++ Java Kotlin TypeScript JavaScript C# HTML/CSS SQL

Machine Learning

PyTorch LangChain (Agents) TensorFlow OpenCV NLP Keras Sci-Kit Learn Matplotlib NumPy / Pandas

Infrastructure & Tools

Docker Kubernetes AWS Google Cloud (GCP) CUDA Redis Flask Node.js / React PostgreSQL / MongoDB WebSockets Git

Campus Leadership & Clubs

UCLA ACM AI

Projects Officer (Nov 2025 - Present)

Led and taught machine learning concepts and applications to 50+ students across quarterly projects. Developed reference codebases and lectured on:

  • Fine-tuning LLMs using Chess-GPT to play strategic chess via RL.
  • Training image diffusion models using PyTorch & VAEs.
  • Building automated video translation pipelines using speech recognition.

UCLA Unmanned Aerial Systems (UAS)

Computer Vision Team (Oct 2025 - Present)

Engineered CV codebases for automated hardware tracking:

  • Designed and manufactured a physical autonomous quadcopter drone.
  • Built the object-tracking computer vision pipeline for an autonomous drone using YOLO and OpenCV.
  • Integrated custom PID control loops inside ROS (Robot Operating System) to target and track humans dynamically in real-time at 50 FPS.

Awards & Honors

Track Winner – LA Hacks 2026

Awarded 1st place out of 76 project submissions in Voice AI & Synthesis (sponsored by ElevenLabs).

3rd Place – UC Berkeley AI Autonomous Racing

Achieved podium placement in the simulation racing series competing against worldwide engineers.

USACO Gold Division Competitor

Ranked in the top 5% of all USA Computing Olympiad competitors, demonstrating proficiency in data structures and advanced algorithms.

National Merit Finalist & UCLA Scholar

Recognized in top 0.1% of graduating US high school seniors; recipient of the prestigious UCLA Engineering Scholarship.