SOFTWARE ENGINEER

DEV JOSHI

DEV JOSHI

I'm a Software Engineering Student at Ontario Tech University, focused on utilizing artificial intelligence and designing pipelines that help create an impact.

View Work
Toronto · Remote Software Engineer Machine Learning Cloud Computing
Scroll
About

About Me

I am currently an AI/ML Engineer at Brilliant Catalyst and an ML Researcher at Ontario Tech University, where I architect self-hosted LLM infrastructure and optimize medical image segmentation models. Previously, I built real-time, high-concurrency systems as a Software Engineering Intern at Intelligent Solutions Lab.

Here are some technologies I am using

  • Python
  • TypeScript
  • PyTorch
  • Next.js
  • Node.js
  • Docker
0Years Experience
0Projects
0Technologies
Experience

Growing Through Impactful Roles

A timeline of the work experiences that accelerated my growth as a software engineering student.

Brilliant Catalyst WIL

AI/ML Engineer

Oct 2025 – Dec 2025 (Part-time Contract)

  • Eliminated recurring AI inference costs and improved response latency by 30% by architecting a self-hosted Ollama LLM infrastructure on AWS EC2
  • Boosted student engagement by engineering a Socratic teaching agent with LangChain and vector-embedded RAG to provide contextual code feedback
  • Maintained 500+ active concurrent coding sessions with minimal latency by building a production-grade online IDE using NextJS, WebSocket connections, and the Monaco Editor
AWS EC2
Ollama
LangChain
Next.js
TypeScript
Monaco Editor
SQLAlchemy
Pydantic

Ontario Tech University

Machine Learning Researcher

Apr 2025 – Nov 2025 (Part-time Research Assistant)

  • Accelerated medical image segmentation model training speeds by 40% by implementing Mixed Precision training on a custom PyTorch U-Net architecture
  • Achieved 93% diagnostic accuracy on 10,000+ MRI scan images by designing a custom weighted loss function to address severe dataset class imbalances
  • Accelerated clinical review workflows by developing an interactive 3D visualization pipeline using Marching Cubes algorithms
PyTorch
AMP
U‑Net
torch.compile
Marching Cubes

Intelligent Solutions Lab

Software Engineering Intern

Apr 2025 – Jul 2025

  • Eliminated manual data lookups for clinical staff and achieved sub-second latency by engineering a real-time Flutter dashboard with WebSocket data streams
  • Maintained API response times under 200ms under peak load by architecting highly optimized NodeJS and Firestore backend connections for 10,000+ users
Flutter
WebSocket
Node.js
Firestore
Jest
Cypress
GitHub Actions
Redis

Code Ninjas

Software Programming Instructor

Feb 2022 – Feb 2024

  • Fostered technical literacy for 100+ students ages 7-14 by breaking down abstract logical reasoning concepts into highly tailored programming lessons
  • Ensured consistent instructional quality and strong parental trust by organizing class schedules and conducting structured progress evaluations
JavaScript
Game Development
Teaching
Featured Projects

Driven by curiosity and continuous improvement.

A curated exhibition of passion projects that improved my foundation and sharpened the blade of my skills.

FeaturedStudio

VeridianAI - Personalized Newsletter SaaS

Built a personalized newsletter SaaS that aggregates and summarizes over 10k daily articles using Hugging Face Transformers, delivering a lightweight UI with Firebase auth and achieving a 95+ Lighthouse performance score –

Python
Firebase
Next.js
Tailwind CSS
    Live Demo
    GitHub
    More Projects
    Project

    StellarSphere - Decentralized FinTech dApp

    Built a decentralized FinTech dApp with a three‑tier microservices architecture, enabling high‑concurrency escrow payments and secure blockchain operations on the Stellar network;

      Node.js
      PostgreSQL
      Docker
      Flask
      Live Demo
      GitHub
      Project

      Censor-It

      A privacy‑focused moderation platform that uses ML models to detect and automatically censor inappropriate text and images in real time, supporting multi‑language content;

        Python
        TensorFlow
        FastAPI
        Docker
        AWS Lambda
        Live Demo
        GitHub
        Tech Stack

        Tools, systems, and platforms I build with.

        A collection of programming languages, machine learning frameworks, and cloud systems I rely on to solve problems and learn.

        Frontend

        Next.js
        React
        TypeScript
        Framer Motion
        Tailwind CSS

        Backend

        Node.js
        Postgres
        GraphQL
        tRPC
        Prisma

        AI/ML

        Python
        TensorFlow
        LangChain
        Vertex AI
        Weights & Biases

        Cloud

        Vercel
        GCP
        AWS
        Cloudflare
        Docker

        Design

        Figma
        Framer
        ProtoPie
        After Effects
        Spline
        Testimonials

        Guided by friends and mentors

        A few words from collaborators and mentors who have supported my growth and shared in my journey.

        “Dev worked with me as part of Brilliant Catalyst’s Work-Integrated-Learning Program and quickly demonstrated an excellent ability to learn new skills, including applying LLMs such as phi-4-mini for AI-assistant coding. Dev showed a unique balance of academic discipline and practical SE, approaching unfamiliar systems with curiosity and follow-through.”

        Qusay H. Mahmoud

        Professor of Software Engineering | AI & Software Innovator