Innovative Projects Realized

Explore thousands of successful projects resulting from collaboration between organizations and post-secondary talent.

30508 Completed Projects

2882
AB
5105
BC
825
MB
681
NL
860
SK
9051
ON
9491
QC
97
PE
586
NB
1141
NS

Projects by Category

Understanding and Improving the Reasoning Capabilities in Vision-Language Models

Vision-language models (VLMs) are considered augmentations of text-only large language models (LLMs) with additional access to the visual world. However, it is known that VLMs typically perform worse than text-only LLMs on reasoning tasks that involve only text (e.g., math word problems), raising the question of whether the benefits of visual augmentation justify the loss in reasoning capabilities. In this project, we will systematically benchmark and analyze the reasoning performance drop of VLMs and propose technologies to improve reasoning in VLMs through both performance-driven analysis and interpreting the internal hidden states of the models.

View Full Project Description
Faculty Supervisor:

Freda Shi

Student:

Partner:

Layer 6 AI

Discipline:

Computer science

Sector:

Finance and Insurance; Professional, scientific and technical services

University:

University of Waterloo

Program:

Accelerate

Designing high precision pointing systems for balloon-borne telescopes

Stratospheric balloons operate above more than 99% of the Earth’s atmosphere, allowing astronomers to observe colours of light that would otherwise be blocked or blurred by the atmosphere at only a tiny fraction of the cost of an equivalent space telescope. However, it is very challenging to ensure that to telescopes are pointing the right location on the sky when they are suspended from a stratospheric balloon.

In this project the student intern will work with StarSpec Technologies to build and test a pointing system for the Gigapixel-class Balloon-borne Imaging Telescope (GigaBIT), a 1.35 meter telescope that will make extremely large and high resolution images of the sky. The student intern will also use their experience building GigaBIT to design pointing control systems that can be applied to StarSpec’s other client balloon telescopes with a wide variety of pointing requirements.

View Full Project Description
Faculty Supervisor:

Laura Fissel

Student:

Partner:

StarSpec Technologies Inc.

Discipline:

Physics

Sector:

Professional, scientific and technical services

University:

Queen's University

Program:

Business Strategy Internship

Pharmacology of psychoactive substances

This project aims to investigate the pharmacology of norbaeocystin, baeocystin, norpsilocin, and aeruginascin – four compounds found in psychedelic mushrooms. The research will examine how these compounds interact with specific brain receptors – the 5HT-1A, sigma-1 and a-adrenergic receptors – that may contribute to the therapeutic effects. These compounds may have potential for the treatment of depression, anxiety, post-traumatic stress disorder and other conditions.
Laboratory testing includes calcium-based assays and ß-arrestin recruitment tests to determine receptor signalling. Compounds demonstrating efficacy in these assays will then be tested. The elevated plus maze test in mice will provide information about the in vivo behavioral effects.

View Full Project Description
Faculty Supervisor:

Robert Laprairie

Student:

Partner:

Ghent University

Discipline:

Life Sciences

Sector:

Education

University:

University of Saskatchewan

Program:

Globalink Research Award

Promutuel Assurance : Modélisation de la probabilité de perte totale

Promutuel Assurance : Modélisation de la probabilité de perte totale
Principales activités du partenaire :
Fondée au Québec en 1852, Promutuel Assurance est une fière mutuelle, présente et engagée avec coeur auprès de ses membres-assurés, de son personnel et de ses communautés. Approche humaine, solutions adaptées, conseils clairs et personnalisés, Promutuel Assurance met tout en oeuvre pour offrir confiance et satisfaction aux membres-assurés. Promutuel offre entre autres, des assurances Auto, Habitation, Entreprise, et Agricole.
Problématique et avantages escomptés du projet :
À la suite d’un sinistre automobile, les assurés contactent leur assureur afin de rapporter le sinistre et amorcer le processus de réclamation. À ce moment, l’assureur récoltera certaines informations auprès de l’assuré afin de comprendre les circonstances du sinistre et l’étendue des dommages. Ces informations, en plus de celles disponibles au contrat d’assurance, permettront de bien guider l’assuré pour la suite des procédures. Dans la majorité des cas, le véhicule sera envoyé chez un carrossier et un estimateur devra évaluer les coûts de réparation avant que le carrossier ne procède à la réparation. L’assureur paiera généralement aussi les frais de location d’une voiture de remplacement jusqu’à ce que le véhicule soit réparé. Cependant, lorsque les dommages sont tels que le véhicule doit être déclaré perte totale, il est préférable de le déterminer rapidement afin d’éviter des coûts d’entreposage du véhicule, de location de voiture, d’estimation et dans certains cas de remorquage additionnels. Un règlement rapide du sinistre permet également d’améliorer l’expérience client. Ce projet visera donc le développement d’un modèle prédictif pour la probabilité de perte totale.
Selon les processus actuels de Promutuel, un questionnaire de 17 questions peut être rempli afin de déterminer si un véhicule est « probablement une perte totale ». Des points sont assignés selon les réponses aux différentes questions. Au-delà d’un certain pointage, un indicateur catégorisera le véhicule comme « probablement une perte totale ».
Promutuel a identifié des enjeux d’affaires avec ce processus et souhaite les résoudre grâce à ce projet. Voici ces enjeux :
• Le questionnaire comporte un grand nombre de questions, ce qui ralentit significativement le processus de réclamation.
• Les questions et les pointages sont basés sur des règles d’affaires et non sur un modèle prédictif.
• Le questionnaire est facultatif. Il est donc rempli seulement si l’expert en sinistre le juge pertinent. Ceci fait vraisemblablement en sorte que plusieurs pertes totales ne sont pas repérées rapidement.
Le projet de stage propose de remplacer ce questionnaire par un modèle d’intelligence artificielle qui utilisera l’information déjà disponible au moment de la déclaration de sinistre afin de prédire la probabilité de perte totale.

View Full Project Description
Faculty Supervisor:

Christian Gagné;Luc Lamontagne

Student:

Partner:

Promutuel Assurance

Discipline:

Computer science

Sector:

Finance and Insurance

University:

Université Laval

Program:

Accelerate

Révolution jumelage : Recherche d’augmentation de données contextuelles pour la prédiction de profils de personnalité des clients et des intentions d’achat par IA

Principales activités du partenaire :
L’entreprise offre des services de développement et d’exploitation de solutions informatiques.

Problématique et avantages escomptés du projet :
Le but de ce projet est d’inférer un profil psychométrique (OCEAN) à partir de questionnaires clients n’incluant pas de questions explicitement psychométriques. On recherchera également des variables contextuelles pertinentes pour enrichir les données multimodales déjà disponibles, afin d’obtenir les meilleures prédictions possibles dans cette tâche de régression multivariée. À terme, cela permettra de mieux comprendre le comportement des clients.

View Full Project Description
Faculty Supervisor:

Christian Gagné;Sebastien Tremblay

Student:

Partner:

Pairing Revolution Inc

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

Université Laval

Program:

Accelerate

Bovine Oligosaccharides and their Effects on Neonatal Calf Gut Function, Microbial Composition, and Development

Newborn calf health is a major challenge, with digestive disorders being one of the main causes of illness, leading to antimicrobial use in the first week of life. While colostrum provides immune factors needed to achieve passive immunity, it also contains prebiotic sugars called oligosaccharides (OS). These sugars are under researched in calves but are shown to improve gut function in human infants as OS feed beneficial bacteria (Bifidobacterium) within the gut. We will test OS supplied by Denkavit to determine what the optimal dose is to stimulate Bifidobacteria growth isolated from neonatal calves. Results will allow Denkavit and its Canadian partner Grober Nutrition advance their understanding of their additive products to potentially add to milk replacer to help improve gut function in calves. This work will overall support the development of more advanced feeding strategies, reduce antimicrobial use, and will help bring innovative, research-driven products to the Canadian and global dairy markets.

View Full Project Description
Faculty Supervisor:

Michael Steele

Student:

Partner:

Denkavit;Grober Nutrition Ltd.

Discipline:

Life Sciences

Sector:

Agriculture

University:

University of Guelph

Program:

Accelerate

Outils d’évaluation ESG pour les projets numériques

Le projet vise à initier le développement d’un cadre méthodologique ESG que les organisations pourront utiliser pour mieux évaluer l’impact environnemental, éthique et social de leurs projets numériques, et qui pourrait être intégré dans l’outil d’évaluation des projets numériques de Prompt.

View Full Project Description
Faculty Supervisor:

Martin Noel

Student:

Partner:

Prompt

Discipline:

Computer science

Sector:

Information and cultural industries; Professional, scientific and technical services

University:

Université TÉLUQ

Program:

Business Strategy Internship

Breast density measurement tool

Lululemon develops and markets athletic apparel including women’s bras. Some women experience difficulty finding the appropriate bra to achieve adequate support without discomfort [1]. Lululemon is experienced in tissue biomechanics models of the breast including measurement of tissue movement, forces and accelerations in an effort to better understand the complex issues around bra performance. Nevertheless, gaps remain in this understanding, and current measurement tools do not provide measurements of all tissue properties, such as breast density. Therefore, opportunities exist to keep improving bra design and fitting. UBC has expertise and novel technology in the measurement of tissue biomechanics including ultrasound-based measurement tools. Ultrasound has been used successfully in women’s health, for almost a century, such as during pregnancy, because of the safe nature of the ultrasound waves and the rich information in the ultrasound signals. This project aims to develop a portable breast density measurement tool that is accurate and easy to use. Such a tool will be used by Lululemon in fundamental research on tissue biomechanics such as enhancing their computational models of the breast. In the longer term, such a tool can be envisioned for use in customized/personal fitting and for adaptive apparel.

View Full Project Description
Faculty Supervisor:

Robert Rohling

Student:

Partner:

Lululemon

Discipline:

Physics

Sector:

Manufacturing; Retail trade

University:

The University of British Columbia

Program:

Accelerate

Leveraging Agentic AI to Detect and Mitigate Shilling Attacks for Enhanced Robustness in Recommendation Systems

This research project explores how agentic AI—autonomous AI agents with the ability to perceive, reason, and act—can be designed and deployed within recommendation systems to proactively identify and counteract shilling attacks. These attacks involve malicious users injecting fake reviews or ratings to manipulate recommendations. The project aims to develop an adaptive, multi-agent, Agentic AI-driven model that monitors user behavior patterns, detects suspicious activities, and dynamically adjusts recommendation outputs to maintain system integrity and user trust.

View Full Project Description
Faculty Supervisor:

Rasha Kashef

Student:

Partner:

Nile University

Discipline:

Computer science

Sector:

Education

University:

Toronto Metropolitan University

Program:

Globalink Research Award

Développement d’un LLM capable de générer automatiquement du texte sous contraintes à des fin d’utilisation en orthophonie

La mission de PhonIA est d’aider les écoles, les orthopédagogues, les orthophonistes et les enseignants à mieux accompagner les élèves présentant des difficultés d’apprentissage. PhonIA a donc développé des outils automatisés de dépistage, d’évaluation et de suivi des difficultés d’apprentissage disponibles via une plateforme numérique. Ces outils permettent d’optimiser et d’améliorer les services professionnels. Les tâches fastidieuses comme: la
retranscription phonétique; l’analyse des erreurs; le suivi des performances; la formulation de recommandations sous forme de rapport sont ainsi simplifiées ce qui permet aux professionnels de se consacrer à des tâches nécessitant leur expertise et à haute valeur ajoutée. En plus de ces outils, PhonIA développe aussi des outils permettant la génération d’activités d’intervention spécifiques et individualisées sur la base des besoins identifiés et documentés grâce à ses outils d’aide à la décision pédagogique et clinique. Dans le futur, PhonIA planifie d’adapter son outil à divers cas d’usage dans le domaine de la cognition, du langage, de la parole, de la phonation, de la communication et des apprentissages.

View Full Project Description
Faculty Supervisor:

Gilles Pesant

Student:

Partner:

Les Solutions PhonIA

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

Polytechnique Montréal

Program:

Accelerate

Quantum Simulations with Physics-Informed Neural Networks

Physics-Informed Neural Networks (PINNs) have emerged as a powerful tool for solving PDEs, fractional equations, and eigenvalue problems, making them particularly suited for quantum mechanics, where traditional numerical methods often face computational bottlenecks. The aim of this internship is to analyse the applicability of PINNs to a broad range of quantum simulation tasks such as solving the Schrödinger equation for multi-electron systems or simulating molecular dynamics for processes like silicidation. The intern will engage in hands-on scientific machine learning, contribute to improvements in quantum simulation, and acquire valuable skills in a rapidly evolving field.

View Full Project Description
Faculty Supervisor:

David Cooke

Student:

Partner:

Fraunhofer-Institut für Integrierte Systeme und Bauelementetechnologie

Discipline:

Physics

Sector:

Quantum Science; Technology

University:

McGill University

Program:

Globalink Research Award

Developing open source user tools for Planet’s hyperspectral remote sensing data

This project will help make it easier for scientists and data engineers to work with a powerful type of satellite data called hyperspectral imagery, which captures detailed information about the Earth’s surface. While this data can support important work in areas like climate, agriculture, and environmental monitoring, it’s often difficult to use without specialized training or tools. The intern will evaluate a range of existing open-source software tools used to process and analyze hyperspectral data, comparing their performance, ease of use, and technical features. This research will identify common challenges and recommend improvements that can make these tools more accessible to a wider community of users.
The project directly supports Planet’s efforts to improve access to data from its new hyperspectral satellite, Tanager-1, by guiding the development of a user-friendly, open-source Python library. The intern’s work will help Planet expand its customer base, support applied research, and contribute to broader goals in open science and environmental technology innovation.

View Full Project Description
Faculty Supervisor:

Nasem Badreldin

Student:

Partner:

Planet Labs Geomatics Corp

Discipline:

Computer science

Sector:

Professional, scientific and technical services

University:

University of Manitoba

Program:

Accelerate