The 5th annual CREATE SE4AI Retreat was held on Saturday, July 4, 2026, at the Concordia University Conference Centre, bringing together software engineering graduate students from Concordia University, Polytechnique Montréal, and Queen's University for a full day of learning, collaboration, and networking.
The retreat began with a keynote presentation by Prof. Benoit Baudry on Dependency Hell Redux, followed by program updates from the CREATE SE4AI leadership team. Participants then took part in the CREATE-a-Thon, a hackathon-style activity where student teams developed proof-of-concept solutions to real community challenges using agentic AI. The event showcased the creativity and technical skills of the participating teams, with solutions addressing topics such as food waste reduction and care coordination for seniors.
The day also included opportunities to connect with fellow trainees and invited guests from industry and academia.
CREATE-a-Thon Projects
FoodBridge
Food Waste Redistribution
Mohamad Salim, Kawsar Bhuiyan, Sharon Ho, Patrick Foalem, Adekunle Ajibode
FoodBridge is a proof-of-concept agentic AI system that tackles the problem of food waste by connecting surplus food from restaurants and grocers with community kitchens and shelters.
View GitHub →CareWithYou: Human-Centered Support for Independent Aging
Senior Care Coordination
Jasmine Latendresse, Yasmine Drissi, Jean Carlors Paul, Sylvain Kouemo, Yu Shi
CareWithYou uses multiple AI agents to coordinate structured senior care through assessing medical needs, safety risks and lifestyle habits. CareWithYou supports care planning without replacing clinical judgement, using only necessary patient information and protecting sensitive medical records.
View GitHub →EzAccess
Accessibility Assistant for Public Services
Samuel Abedu, Gideon Peters, Anthony Forgetta, Yvan-Marcel Carré, Mahi Begoug
EzAccess presents a proof-of-concept agentic AI system that connects users to over 2,400 public places in Montreal such as parks, libraries, community centres, etc.
View GitHub →AEGIS: Agentic Emergency Guidance and Incident Support
Climate Emergency - Flood Response Assistant
Alor Ebube, Ghada Abushaqra, Khandaker Rifah Tasnia, Yasmine Ben Braiek, Aroh Sunday Melitus
AEGIS is a working proof-of-concept (POC) agentic solution to help flood-affected individuals request urgent aid, find safe shelters, and connect with local volunteers and NGOs in real time. AEGIS converts fragmented reports into prioritized requests, verified resource matches, live routes, public guidance and an auditable human authorized response.
View GitHub →Job-Fit Gap Analyzer
Job Search and Career Transition Support
Musfiqur Rahman, Rachna Raj, Fazle Rabbi, Minh Nguyen
The Job-Fit Gap Analyzer is a static browser-based tool that evaluates a user's career fit by comparing their CV against aggregated real-world job posting data. It allows users to benchmark themselves against general market trends, analyze their suitability for specific job postings, or utilize a "Career Pathfinder" to find optimal transition routes between different roles. All analysis, including CV parsing and skill matching, runs entirely client-side, ensuring privacy without the need for a backend server. The tool relies on data compiled from Hugging Face datasets, which users can regenerate or customize using provided Python scripts. Ultimately, it provides actionable insights into skill gaps, experience alignment, and market demand for six core career paths.
View GitHub →Smart Schedule Health
Medical Appointment Coordination
Antonio Collante, Lucas Armand, Junjie Li, Aurel Honey, Gen Caumartin
An agentic AI proof-of-concept solution that uses on-device AI agents to help patients and caregivers organize, reuse, and streamline their information across the medical scheduling process, reducing the repetitive and stressful back-and-forth of booking appointments while keeping sensitive data local and humans in control.
View GitHub →
