About me

Welcome to my homepage! I'm Ayush Kumar, a passionate software engineer and proud recipient of a 100% scholarship at Arizona State University, where I’m pursuing my Master's in Computer Engineering. With a strong academic foundation and hands-on experience, I’m driven to deliver innovative, high-performance solutions.

I specialize in:

  • - Full-Stack Development: Crafting dynamic applications with frameworks like React.js, Node.js, and Next.js.
  • - Cloud and Microservices: Building robust, scalable architectures using AWS, Docker, and Kubernetes.
  • - Data Pipelines and Analytics: Designing ETL pipelines and leveraging tools like Python, Apache Kafka, and TensorFlow for insightful analytics and AI-powered integrations.
  • - Enhanced User Experiences: Optimizing performance, SEO, and server-side rendering for seamless and engaging user interfaces.

Some of my achievements include developing Connectify, a real-time chat application handling 10,000+ concurrent users, and Virtuous Vehicles, an AI-powered vehicle behavior mapping system that secured $300,000 in federal funding.

Explore my portfolio to see more of my work, and let’s connect to discuss how I can help create impactful and innovative software solutions!

What Am I Skilled At?

  • design icon

    Programming Languages

    Python, C++, C#, JavaScript, TypeScript, SQL, Bash

  • frontend icon

    Frontend / Web Frameworks

    ReactJS, Vue.js, Angular, HTML5, CSS3, Tailwind, Bootstrap, jQuery

  • backend icon

    Backend / API Development

    REST APIs, .NET Core, Node.js, Express, FastAPI, Django, MVC architecture, Azure Functions

  • cloud devops icon

    Cloud & DevOps

    Azure DevOps, Docker, GitHub Actions, Git, Jira, Agile (Scrum, Kanban), CI/CD pipelines

  • database icon

    Databases / Data Modeling

    SQL Server, PostgreSQL, MySQL, MongoDB, ORM tools (EF Core, Prisma, SQLAlchemy)

    • Experience

    Arizona State University

    1. Software Engineer

      May 2024 – Present | Tempe, AZ

      1. Building three scalable full-stack platforms serving over 90,000 users, allowing orchestration of AI features using Large Language Models (LLMs) via LangChain, with support for audio, and text inputs.

      2. Developed 80+ reusable components for company’s web applications using React.js and Typescript, translated from Figma designs resulting to a scalable flow for 90,000+ users per month across multiple devices and landing pages.

      3. Designed Java Spring Boot microservices with PostgreSQL and REST APIs, boosting performance and driving a net gain of 4,000 organic student acquisitions annually.

      4. Collaborated with a team of 8 people for a $2M U.S. Government project (Link), and built custom functions for required functionality and integrated third-party databases using REST APIs for data-acquisition.

      5. Improved database indexing and wrote SQL queries, boosting performance for data retrieval in analytics and reporting by 40%.

      6. Configured CI/CD pipelines to build Docker images, deploy serverless REST APIs as AWS Lambda functions, and manage workloads on Kubernetes (Amazon EKS), reducing deployment time by 40% and improving system scalability.

    Autonomous Systems Lab (ASU)

    1. AI Software Developer

      Dec 2023 – Feb 2024 | Tempe, AZ

      1. Fine-tuned a disease-specific LLM (Medgemma 4B IT) using Pytorch, Pinecone, and QLoRA for binary and multi-class detection of diabetes, improving classification precision from 71% to 87% through multi-domain generalization.

      2. Built an internal monitoring dashboard (Grafana, Prometheus) to track GPU/CPU usage, error logs, and memory, enabling smooth research experiments on 10k+ diabetic retinopathy images for Mayo Clinic.

    Cutso Foods LLP

    1. Software Engineer Intern

      May 2022 – May 2023 | Bangalore, IND

      1. Revamped the frontend with a decoupled JavaScript architecture and dynamic filtering features, reducing load time by 1.8s and boosting Lighthouse mobile score from 62 to 91, enhancing user experience on low-bandwidth devices.

      2. Automated Jenkins CI/CD pipeline with Dockerized services, branch-level preview deployments, and schema validation, reducing release regressions by 30% and ensuring deployment readiness.

      3. Standardized responsive layouts (HTML5/CSS3), by resolving cross-device issues & raising order completion rates by 18%.

    Education

    1. Arizona State University

      Master of Science (MS) Computer Engineering (Computer Systems) 2023 — 2025 GPA: 3.9/4.0

      Coursework- Data Structures and Algorithms, Artificial Intelligence, Data Visualization, Blockchain


    2. Dr. APJ Abdul Kalam Technical University

      Bachelor of Technology (B. Tech) Electronics and Communication Engineering 2019 — 2023 GPA: 3.52/4.0

      (1st Rank Holder in the batch of 1100 students)

      Coursework- Data Structures, Object Oriented Programming, Computer Networks


    3. Central Board of Secondary Education

      Secondary School Certificate (SSC) Science with Mathematics March 2018

    Projects

  • portfolio icon

    AI-Powered Portfolio Website

    - Built an interactive portfolio with a RAG-based chatbot (Next.js, TypeScript, Tailwind).
    - Enabled visitors to query resume, education, and project details dynamically.
    - Designed to provide a personalized, AI-driven browsing experience.

    View Project
  • design icon

    RepGenie

    - Engineered the backend of an AI-powered fitness coaching platform using Java and PostgreSQL.
    - Integrated OpenAI APIs for personalized training plans tailored to individual user needs.
    - Deployed on AWS cloud-hosted services, achieving scalability and reliability for 1,000+ active users.
    - Leveraged TypeScript for API design and enhanced type safety, ensuring fewer runtime errors.

    View Project
  • design icon

    Connectify

    - Built a real-time chat application with Next.js, WebSockets, and ConvexDB, scaling to handle 10,000+ concurrent users.
    - Developed scalable REST APIs using Node.js and Express.js, ensuring 99.99% uptime for messaging services.
    - Integrated user authentication with Clerk, JWT, and OAuth2, improving onboarding efficiency by 30%.
    - Implemented React Dropzone and AWS S3 for file uploads, reducing storage errors by 20%.
    - Integrated OpenAI GPT-4 with TypeScript for LLM-powered messaging, reducing response time by 40%.
    - Designed WebSocket-based message queues and user presence tracking in Python backend, supporting 500+ concurrent users with low-latency delivery.

    View Project
  • design icon

    Virtuous Vehicles (Mapping Human Driving Behaviour in Autonomous Cars using A.I.)

    - Engineered AI-enabled vehicle behavior mapping with Python, CARLA Simulator, & OpenCV, boosting accuracy by 35%.
    - Developed a React.js dashboard with Flask APIs, cutting analysis time by 40% and boosting performance by 50%.
    - Automated deployment and scaling with Docker and Kubernetes, ensuring fault tolerance and 99.9% uptime.
    - Designed a backend system using PostgreSQL and TensorFlow for data analysis, securing $300,000 in federal funding.
    - Integrated TensorFlow for object detection, improving precision and safety metrics by 30%.

  • design icon

    COVID Management System (Vaccination Data Management made easy)

    - Designed a COVID-19 data platform using React.js, Node.js, and Bootstrap, efficiently serving 10,000+ users.
    - Created RESTful APIs with Express.js, halving scheduling time and improving response times by 40%.
    - Ensured data integrity and scalability using MongoDB and Redis, maintaining 99.9% uptime.
    - Automated CI/CD workflows with Docker and GitHub Actions, cutting deployment time by 25%.
    - Enhanced system observability with ELK Stack, reducing debugging and resolution times by 30%.

    • Daniel lewis

    Daniel lewis

    Contact

    Contact Form