Hey, I'm Umair

CS · Software Engineering · ML

A third-year computer science student at the University of Toronto , specializing in Artifical Intelligence.

Contact me Download Resume
Scroll down

About Me

I'm a Computer science student at the University of Toronto with a passion for machine learning and software engineering. I enjoy building applications that solve real-world problems and make everyday life better. With experience in collaborative and fast-paced environments, I thrive in teams that value innovation, creativity, and impact. I'm always looking for opportunities to learn, grow, and contribute to meaningful projects.

3+ Full Stack Projects
Shipped
250+ Yearly Github
Contributions
4 Months
Work Exp.

Skills

My tech stack

Languages

  •  Python
  •   Java
  •   C
  •   JavaScript
  •   SQL

Tools

  •   Git
  •   Docker
  •   Firebase
  •   Azure

Frameworks

  •   React
  •   Tailwind
  •   Django
  •   Flask
  •   Pandas
  •   Sci-kit Learn
  •   JUnit

Experience

My journey
Professional Journey

University of Toronto

Toronto, ON
BSc, Computer Science - AI & Systems
2023 - Present
View More

Coursework:

  • Data Structures
  • Software Design
  • Unix Systems
  • Linear Algebra
  • Probability and Statistics

  • Computer Science Major with AI and Systems Specialization

UrbanFood Alliance

Franklin Township, NJ Software Engineer Intern (ML)
Jun 2025 - Aug 2025
View More

Description:

  • • Led development of an internal AI agent using the MCP framework, serving 10+ teams across the organization
    • Boosted response quality by 27% through supervised fine-tuning and iterative feedback-driven model iteration
    • Engineered a RAG pipeline using llama index to vectorize internal data, improving LLM response relevance by 25%
    • Enabled real-time business insights by integrating 12+ APIs and web scraping pipelines across core functions
    • Collaborated with 10+ teams in an Agile environment, using CI/CD pipelines to ship updates 40% faster

  • https://www.urbanfoodalliance.org/

Department of CS, UofT

Toronto, ON Software Developer (Open-Source)
June 2025 - Present
View More

Coursework:

  • Optimizing core features of PythonTA, a static analysis tool used by 4,000+ students annually to improve coding practices in introductory Python courses at the University of Toronto.
    https://github.com/pyta-uoft/pyta/

Projects

Check out some of my recent work!

NxtMovie

NxtMovie is an interactive movie recommendation app developed using Python, Streamlit, and scikit-learn. It employs machine learning algorithms, decision trees, and graph algorithms to provide personalized movie suggestions. The app also uses Beautiful Soup for web scraping real time data and Firebase for secure authentication and data storage. NxtMovie boasts a tailored and secure user experience in an intuitive, user centric interface. Try it out yourself or watch a demo using the links below!

GitHub Repository Try it Out

PromptLink

PromptLink is a full-stack AI platform that intelligently detects user intent and dynamically routes prompts to the most suitable LLM (e.g., GPT-4o, Gemini, GPT-3.5) based on task complexity and context. Built to improve transparency, performance, and cost-efficiency in LLM-powered apps, the project showcases advanced orchestration, interface design, and reasoning analysis.

GitHub Repository Try it Out

SignSpeak

SignSpeak is a real-time sign language translator application built during Gen AI Genesis 2025, Canada's largest AI hackathon. It leverages machine learning for hand gesture recognition using TensorFlow, OpenCV, and Flask, and integrates a web-based frontend built with HTML, CSS, and JavaScript. The system processes live video input to detect sign language gestures and includes voice-to-text functionality for non-signing users via the Web Speech API, creating a seamless, inclusive communication experience in real-time.

GitHub Repository Devpost

Dr Mario in MIPS Assembly

Dr. Mario is a recreation of the classic 1990 Nintendo game where players eliminate viruses by aligning capsules of matching colors. The project uses Memory Mapped I/O for keyboard input and pixel rendering, ensuring smooth gameplay. Key features include capsule movement, rotation, collision detection, and virus elimination, with advanced additions such as gravity, capsule preview, and score tracking.

GitHub Repository

GitHub

Check out my GitHub page to see more of my work!

Link

Contact Information

Get in touch!

Email

contact@uarham.me

Location

Toronto, Ontario, Canada