Hi, I'm
Ryan Merheby
Software Engineer & CS Student · Prev @ Meta
HBSc Computer Science at the University of Toronto.
Prev AI Infra Intern at Meta.
01. About Me
I study Computer Science at the University of Toronto graduating Apr 2026, recognised as a Dean's List Scholar and Chancellor's Scholarship recipient.
I've had 20+ months of engineering experience across three internships — most recently as an AI Infrastructure Intern at Meta, where I worked on model processing pipelines for Instagram & Ads, used by ML engineers 15K+ times daily.
My interests are in AI infrastructure, distributed systems, and full-stack development. Outside of industry I've done Undergraduate Research in generative AI and taught software design as a Teaching Assistant for CSC207.
- Python
- Java
- C++
- JavaScript
- SQL
- Haskell
- React
- Django
- PyTorch
- Git
- Groovy
- HTML / CSS
02. Experience
Teaching Assistant
University of Toronto · Toronto, ON, Canada
- Leading weekly labs for CSC207 (Software Design) in the Fall semester.
- Grading midterms, exams, and final projects.
Software Engineer Intern — AI Infrastructure
- Used PyTorch to optimize model processing pipelines for Instagram & Ads at scale.
- Built a multi-stage validation component in the model processing workflow, closing a validation gap that impacted many production models and caused major outages.
- Extended the project by creating a suite of CLI tools for ad-hoc weight validation of published model snapshots.
- Implemented distributed chunk processing of weight pruning, decreasing peak memory usage by 14%.
Software Engineer Intern
Veeva Systems · Toronto, ON, Canada
- Led development of a Java internal debugging tool allowing secure API calls in customer instances.
- Added multivalue fields for user info & optimized aggregation calls, reducing user table load times by ~0.2 seconds.
- Created new DB tables, service & API for a new revision history feature for medical data processing tasks.
- Wrote Groovy scripts for Pfizer & Bayer to convert medical data to match regional standards.
Student Researcher
University of Toronto · Toronto, ON, Canada
- Wrote a research report & presented on applying 1st-order logic to create explainable AI in generative design.
- Built a pipeline converting natural language → 1st-order logic → generative design models using the Fusion 360 API and a fine-tuned LLM.
Data Engineer Intern
Careem · Dubai, UAE
- Designed a dynamic resource routing model for PrestoSQL queries, boosting query performance by ~10%.
- Conducted data visualization & built Grafana performance dashboards for the query engine.
- Developed features & wrote JUnit/Mockito tests for the migration from PrestoSQL to Trino.
03. Projects
PetPal
Led a team of 4 developers to build a React & Django pet adoption website. Designed the UI & backend services for user profiles, adoption listings, and pet info pages.
DebrisMine
Environment-focused social media app for non-profit Oceanic Impact. Worked with 7 developers to expand the React Native app with groups, article posting, likes & comments. Implemented Django backend MVC and Selenium article web scraping.