Ayush Desai's Resume

Ayush Desai

Full-stack Software Engineer shipping scalable, production systems across Python, TypeScript, and AWS.

Las Vegas, NV, United States, PST

Ayush Desai's profile picture

About

Full-stack Software Engineer with 2+ years shipping scalable, production systems across Python, TypeScript, and AWS. Owns features end-to-end — from design through deployment — with a focus on performance, reliability, and developer velocity. M.S. Software Engineering candidate seeking full-time SDE roles; authorized to work in the U.S. (OPT, June 2026).

Work Experience

Kinship AI

Jun 2025 - Present

Software Engineer

  • Re-engineered Shopify customer data sync into an async, webhook-driven Cloudflare Workers job, improving sync speed 40x, auto-relinking existing records to Shopify customer IDs on store connection, and enforcing GDPR-compliant handling with 30-day deletion SLAs for imported customer data.
  • Instrumented platform-wide error monitoring and alerting with Sentry, triaging 33+ production issues and sharply reducing time-to-diagnosis for critical bugs.
  • Shipped a scalable Shopify integration that let customers securely import and sync store data into Kinship.
  • Stood up the platform's frontend and backend test suites from scratch, lifting coverage and catching regressions before they reached production.
  • Rebuilt the database migration flow to eliminate deployment failures and enable safe schema changes across every environment.
  • Identified and remediated multiple API-layer security vulnerabilities, hardening the platform against data-exposure risks.
  • Resolved a race condition between database writes and Twilio message insertion, eliminating message loss in customer conversations.
  • Re-architected the metrics layer to isolate analytics reporting per organization across pipelines, restoring multi-tenant data accuracy.
  • In-Person
  • TypeScript
  • Cloudflare Workers
  • Shopify
  • Sentry
  • Twilio

EdPlus at Arizona State University

Mar 2025 - May 2026

Software Developer

  • Built and deployed a Next.js + TypeScript staff evaluation platform for 50+ staff, integrating Google Apps Script for automated data provisioning and Google Gemini for AI-generated summaries.
  • Developed a dockable Unity editor tool backed by Coda REST APIs, letting 30+ developers view, filter, and update 450+ project tasks in-engine and cutting context-switching.
  • Engineered a serverless AWS ETL pipeline (Lambda + S3 + Apps Script) that automated scheduling workflows and powered Looker Studio dashboards for 100+ staff.
  • Built a FastAPI video-compression microservice for VR QA captures, shrinking file sizes 60% on average (up to 89%) inside the automated VR build pipeline.
  • Extended a CLI, REST API, and web UI for VR build distribution, cutting provisioning time 70% and enabling local testing for the internal QA team.
  • Migrated the Dreamscape Learn platform from Vue 2 to Vue 3 (Composition API), improving performance, modularity, and long-term maintainability.
  • Automated reporting with Python ETL scripts, cutting reporting time 60% and increasing throughput across EdPlus VR Pods.
  • Built a Perforce trigger framework enforcing Unity asset integrity (meta-file validation, GUID consistency) with automated file locking, eliminating recurring merge conflicts across the VR build pipeline.
  • Extended a Unity Git/Perforce plugin with a Submit Changelist feature and automated token refresh, eliminating session timeouts and streamlining deployments.
  • Hybrid
  • Next.js
  • TypeScript
  • Python
  • AWS
  • FastAPI
  • Vue 3
  • Unity
  • Perforce

Skitti (IIT Mandi-funded startup)

May 2024 - Aug 2024

Software Engineer

  • Built and shipped a real-time multiplayer quiz module in React Native to the Google Play Store, powering live lobby creation, friend invites, and head-to-head competition.
  • Integrated AWS Lambda and MongoDB APIs to manage sessions, quiz data, and live game state, reducing response latency and smoothing the multiplayer experience.
  • Remote
  • React Native
  • AWS Lambda
  • MongoDB

WebForest LLP

Jan 2024 - May 2024

Software Engineer

  • Built a full-stack MERN social platform with scalable APIs, JWT authentication, and Stripe webhook integration, enabling secure paid verification and premium-account monetization.
  • Developed a modular React.js e-commerce platform on a headless CMS with a Docker-based CI/CD pipeline, cutting deployment time 25%.
  • In-Person
  • React
  • Node.js
  • MongoDB
  • Docker
  • Stripe

Legal Square Tech

Jan 2023 - Dec 2023

Software Engineer

  • Built React.js frontend components against GraphQL and Java Spring Boot REST APIs, keeping the cloud-based legal training platform on schedule for new course launches.
  • Cut post-release bugs by introducing Jest + GitHub Actions automated testing and coordinating defect fixes across frontend and backend layers.
  • Delivered features end-to-end in two-week Agile sprints (JIRA), translating product requirements into technical tasks alongside senior engineers.
  • Remote
  • React
  • GraphQL
  • Java Spring Boot
  • Agile

Education

Arizona State University

Mar 2024 - May 2026
M.S., Computer Software Engineering · Fulton Schools Graduate Scholarship · Engineering Graduate Fellowship

Skills

  • Python
  • TypeScript/JavaScript
  • Java
  • SQL
  • React/Next.js/Vue 3
  • Node.js/Express.js
  • FastAPI/Django
  • REST APIs
  • GraphQL
  • AWS (Lambda, S3, DynamoDB, API Gateway, EC2)
  • Cloudflare Workers
  • Docker
  • GitHub Actions/Bitbucket Pipelines
  • Perforce
  • PostgreSQL/DynamoDB/MongoDB
  • LLMs/RAG/LangChain
  • Vector Databases (Pinecone)
  • Tailwind CSS
  • System Design
  • Microservice Architecture

Side projects

Enhanced Tracking and Reporting of Missing Persons — IEEE COMPSAC 2025

Published at IEEE COMPSAC 2025. Designed a knowledge-graph-based distributed tracking system using GraphDB, SPARQL, and REST APIs, with a React.js frontend and Express.js backend.

  • Python
  • GraphDB
  • SPARQL
  • React.js
  • Express.js
  • Node.js

Mety Legal — Sponsored Capstone Project

Full-stack legal chatbot with a Retrieval-Augmented Generation pipeline — document ingestion, embedding generation, vector search (Pinecone), and LLM orchestration via LangChain and LangGraph — delivering context-aware legal responses.

  • React.js
  • TypeScript
  • Django
  • LangChain
  • LangGraph
  • Pinecone

Convoz

Full-stack social media application built from scratch using the MERN stack with Stripe-powered paid verification.

  • React.js
  • Express.js
  • Node.js
  • Tailwind CSS
  • MongoDB
  • Stripe
  • JWT