Projets novateurs réalisés

Common Risk Factors in the Stablecoin Market

This project focuses on understanding what makes the prices of stablecoins (a type of cryptocurrency designed to stay stable) go up or down. Traditional financial models don’t fully explain how these digital assets behave, especially during market changes or when new regulations are introduced. By studying both general market trends and specific factors like liquidity, issuer trust, and new laws like the MiCA regulation in Europe, this research aims to create a better tool for evaluating stablecoin risks. The expected benefit to the participating institutions is the opportunity to strengthen international research collaboration, share advanced expertise in digital finance, and enhance their reputation as leaders in innovative financial research.

A wireless headset for power-aware EEG/non-EEG signals processing and seizure warning – Year two

Epilepsy is a debilitating neurological disorder affecting approximately 50 million people worldwide (World Health Organization). This project proposes to develop a SMART (Seizure Monitoring At the Right Time) Headset for these patients to monitor their brain signal and track their movements, heart rates and breathing patterns. This headset will have new types of sensors suitable for long term usage and sleep. An intelligent micro-chip in the headset will track patient continuously and detect epilepsy syndromes, and warn patient and caregiver in advance (e.g., few seconds to minutes), which may prevent a patient’s life threatening accident. This SMART Headset will connect globally through patient’s mobile device and allow the clinicians monitor the patient remotely 24-hr/7-day with less clinical expense and government healthcare cost. This device will provide continuous personal assistance and may allow patients to work full-time and live independently, and reduce government burden in healthcare and boost economic growth.

Stage en intelligence artificielle – Optimisation du jumelage talent-opportunité à l’aide de l’intelligence artificielle appliquée au recrutement

Le projet vise à structurer et exploiter la base de données de NexWav afin d’entraîner des modèles prédictifs avancés en intelligence artificielle. Cette initiative permettra de :
• Optimiser le jumelage entre candidats et opportunités : en identifiant les meilleures correspondances entre les profils des talents et les besoins des entreprises, réduisant ainsi le temps de recrutement et augmentant la satisfaction des parties prenantes.
• Anticiper les besoins du marché : en prédisant les tendances en matière d’emploi et les compétences recherchées, permettant à NexWav de conseiller proactivement ses clients et candidats.
• Renforcer la position concurrentielle de NexWav : en offrant des services différenciés basés sur l’IA, l’entreprise se positionne en tant que leader innovant dans le secteur du recrutement au Canada.
• Créer de nouvelles opportunités de carrière : en identifiant des parcours professionnels alternatifs pour les candidats, favorisant ainsi la mobilité et le développement des talents

Adaly.AI Generative AI & Data Science Proposal (OTU)

This project focuses on researching and developing an advanced coordination mechanism between our API layers and LLM infrastructure. The goal is to efficiently access and reason across both structured and unstructured data sources to improve customer’s decision-making processes. Success will be measured by the speed in which that data source can be identified, sourced, and integrated alongside the response quality that’s ultimately generated from the implementation.

Resonateur à ondes acoustiques de surface transversal (SH-SAW) sans réflecteurs pour la caractérisation biologiques

La technologie de biodétection est essentielle pour les diagnostics médicaux, la surveillance de l’environnement et la recherche biologique fondamentale. Alors que la résonance plasmonique de surface (SPR) offre une détection sans étiquette sensible aux changements d’indice de réfraction et que les capteurs à ondes acoustiques de surface (SAW) excellent dans les milieux liquides et sondent les propriétés mécaniques, tous deux se heurtent à des limites en termes de sensibilité ultime et de richesse d’information. L’acousto-plasmonique, qui combine l’acoustique et la nanophotonique, offre une voie transformatrice pour surmonter ces limitations. Ce projet vise à concevoir, fabriquer et valider de nouveaux biocapteurs acousto-plasmoniques intégrés en tirant parti du couplage élasto-optique entre les SAW et les plasmons de surface localisés (LSP) dans des architectures multicouches nanostructurées. En générant des SAW à haute fréquence (par exemple, des SAW horizontaux à cisaillement pour une performance optimale des liquides et un couplage électromécanique) qui modulent dynamiquement les propriétés optiques des nanostructures plasmoniques (via des changements d’indice de réfraction induits par la déformation), nous obtiendrons une sensibilité accrue et permettrons une détection multimodale, en corrélant les changements de signaux optiques et acoustiques lors de la liaison de biomolécules.

Physics-informed machine learning for cryogenics optimization

During this Accelerate internship, we will push ZPC’s current dilution refrigerators to new heights of performance and reliability by harnessing modern machine-learning and AI techniques.

L2M Validation / QC Automne 2025 / AsbabAI est une plateforme novatrice qui évalue la causalité des modèles d’apprentissage profond afin d’améliorer leur explicabilité et leur robustesse en utilisant l’inférence causale, une exigence clé dans les secteurs réglementés comme la santé et la finance

Artificial intelligence (AI) is increasingly used to make high-stakes decisions in fields like finance, insurance, and healthcare. However, most of AI systems function as “black boxes,” producing results without clear explanations. This lack of transparency can lead to serious consequences, such as denying loans to creditworthy individuals or making unfair risk assessments.

The AsbabAI aims to make AI models more transparent, reliable, and robust. Using advanced techniques called causal inference, it helps uncover the real reasons behind a model’s decisions and ensures those decisions remain valid even when real-world conditions change.

This project will be especially valuable for financial institutions, which face growing regulatory pressure to ensure fairness and explainability in their AI systems. It also supports public trust in AI and contributes to Canada’s leadership in responsible AI innovation.

A wireless headset for power-aware EEG/non-EEG signals processing and seizure warning

Epilepsy is a debilitating neurological disorder affecting approximately 50 million people worldwide (World Health Organization). This project proposes to develop a SMART (Seizure Monitoring At the Right Time) Headset for these patients to monitor their brain signal and track their movements, heart rates and breathing patterns. This headset will have new types of sensors suitable for long term usage and sleep. An intelligent micro-chip in the headset will track patient continuously and detect epilepsy syndromes, and warn patient and caregiver in advance (e.g., few seconds to minutes), which may prevent a patient’s life threatening accident. This SMART Headset will connect globally through patient’s mobile device and allow the clinicians monitor the patient remotely 24-hr/7-day with less clinical expense and government healthcare cost. This device will provide continuous personal assistance and may allow patients to work full-time and live independently, and reduce government burden in healthcare and boost economic growth.

L2M – EnerCast

I intend to commercialize a foundation model for forecasting energy production and consumption time series, addressing challenges like renewable integration, storage optimization, and demand-supply balancing. Resources such as HVACs, EVs, ESSs, and solar roofs introduce uncertainty due to temperature sensitivity, price response, and weather variability, making grid management and the transition to smart grids more complex. Our model incorporates external factors like outages and extreme events to improve forecast accuracy and mitigate peak prices. We also aim to build automated energy management systems that reduce customer costs while helping system operators balance the grid. Additionally, we plan to develop an LLM-based chatbot that offers intuitive, AI-driven insights for market participants and energy traders, making advanced forecasting accessible without requiring technical expertise.

L2M – IoSL: A Geospatial-aware Internet of Safe Labourers for Assessing Workers’ Well-being and Safety Improvement

This project will develop a plan to bring a new worker fatigue monitoring system to market. The system uses wearable sensors and artificial intelligence to detect when workers are becoming tired, so that action can be taken to prevent accidents and keep people safe on the job. During this project, market research will be done to understand customer needs, and strategies will be created for testing the system in real workplaces and launching it as a business. The expected benefit to the partner organization is a clear and practical roadmap for successfully commercializing this technology, thereby improving worker safety and productivity in industries such as construction, mining, and transportation.

L2M – Vira

This research-based project aims to make developing virtual reality (VR) applications faster and easier by using generative AI tools. By streamlining the development process, the project can help both companies and individual developers save time and resources. This is especially valuable for Calgary’s growing community of VR businesses and research labs. Overall, the project supports Calgary’s and Canada’s innovation ecosystems by introducing new ways to boost productivity and creativity in VR development.

L2M – Clinical Trial Recommendation System Using Large Language Model

We are developing a conversational AI system that helps hospitals and cancer centers find the right patients for clinical trials much faster and easier. Currently, research staff spend weeks manually reading through hundreds of pages of patient medical records to see if patients qualify for specific cancer treatment studies, which means many patients who could benefit from new treatments are never found or found too late. Our AI agent can automatically read and understand patient records, medical notes, and lab results, then have natural conversations with healthcare workers to recommend which trials might be good matches for each patient. This system will save significant time and money by reducing the manual work needed to screen patients, help more cancer patients access potentially life-saving treatments, and enable smaller hospitals to participate in clinical research that they couldn’t handle before due to limited staff resources. The conversational interface makes the technology easy for busy healthcare professionals to use without extensive training, ultimately improving patient care while reducing administrative burden on medical staff.