Research Assistant 1

McGill University

2d • $30 • Onsite

About The Position

This Part-Time Research Assistant 1 position is based in Prof. Jeremy Cooperstock’s research lab in the department of Electrical and Computer Engineering. Job Summary We are seeking a part-time Research Assistant to take on a research & development role in the Shared Reality Lab’s open source IMAGE project (image.a11y.mcgill.ca). The candidate would be responsible for defining and implementing all AI aspects of IMAGE. This role has two key parts: 1. We are collaborating with Humanware, a leading provider of solutions for people who are blind or have low vision, to design and develop accessible textbook‑diagram experiences for users of the Monarch Android pin‑array tablet (https://www.humanware.com/en-usa/monarch). In this role, you will work closely with Humanware’s ML developers to design and implement an ML toolchain that extracts meaningful information from a wide range of textbook diagrams. The resulting structured outputs will support designers in creating high‑quality tactile representations of these diagrams for blind and low-vision learners. 2. The IMAGE architecture was originally designed in 2021, before the widespread use of LLMs. We have moved to LLMs for much of the technical stack, but now we need to work towards deeper integration. Using your knowledge of the current AI landscape and likely future directions, you will propose and implement changes to the IMAGE architecture and existing components. The goal is to make IMAGE the platform of choice for external developers to convert visual media to audio/haptic experiences that enrich understanding for people who are blind or low vision.

Requirements

• Quebec resident (mandatory), and able to work on-site at McGill downtown Montreal campus at least one day/week.
• Linux command line
• Familiar with (or able to quickly learn) Docker, containerization, client/server design
• Git, and Github flow method of raising and merging pull requests
• Python
• JSON schema definition and use
• Mentoring junior designers/developers
• Fluent English
• Bachelor’s degree in a related field.
• Minimum 2 years of relevant experience.

Nice To Haves

• Experience writing conference and journal papers and presenting results in academic venues
• French, spoken and/or written
• Experience working with the blind/low vision community

Responsibilities

• Conduct experimentation with techniques such as RAG, few‑shot training, and related methods to reliably extract key information from textbook diagrams and other visual media including photos, maps, and graphs.
• Execute informed tradeoffs in accuracy, performance, and cost across model families, model sizes, and inference settings (local vs. cloud).
• Review and revise existing IMAGE machine‑learning components to optimize efficiency, scalability, and quality of results.
• Implement conversions of existing non‑LLM components to LLM-based approaches where appropriate.
• Design and implement modifications to the core IMAGE architecture to improve usability for external development teams, including restructuring components that may align with MCP‑style patterns.
• Prepare and document example code, usage guides, and reference materials to support external users building solutions on the IMAGE platform.
• Review and merge code contributions from students and other team members to ensure consistency and quality.
• Analyze and interpret research outcomes from system improvements and contribute to academic publications derived from project work.

What This Job Offers

• Job Type: Part-time
• Career Level: Entry Level