About
I am a PhD student in Software Engineering at McMaster University, supervised by Dr. Sébastien Mosser under the FATES-MLOps project. My research sits at the intersection of AI safety, assurance case methodology, and MLOps — developing lightweight, verifiable approaches to validate and certify the requirements of AI and machine learning systems.
I have published at venues including ICSE-NIER, ECMFA and MoDRE, with work spanning model card evaluation, safety modelling for ML pipelines, and fairness justification diagrams. I also hold a Bachelor of Engineering in Software Engineering from McGill University.
Publications
MODELS ↗ · Doctoral Symposium Paper
When Safety Claims Fail Silently: A Modelling Approach to AI Assurance
Proposed a modelling language for AI safety claims that turns natural-language model and system cards into structured justification diagrams, exposing where safety arguments break down across the AI lifecycle.
Oct 2026
MODELS ↗ · Tutorial
Modelling Justifications to Support and Operationalize Machine Learning Pipelines
Co-authored a tutorial where we modelled and validated ML pipeline design rationale using jPipe to combat "decision drift", integrating automated conformance checks into CI/CD for reproducible, requirement-traceable pipelines.
Oct 2026
MoDRE ↗ · Workshop Paper
Safety First! Modelling Requirements from GPT-5 System Card using Lightweight Safety Models
Lead-authored a paper where we performed an in-depth manual analysis of the 60-page GPT-5 system card, eliciting 74 claims and 22 implicit safety requirements, classified against the MIT AI Risk Repository taxonomy.
Analysis reveals uneven, tightly coupled, and under-specified requirement coverage across privacy, safety, and misinformation domains.
Applied justification diagrams to formally model requirement compliance (e.g., chemical/biological capability thresholds), surfacing OpenAI's heavy reliance on safeguarding, benchmarks, and external red-teaming as evidence.
Aug 2026
RE ↗ · Tutorial
Validating Requirements of Machine Learning Pipelines with Justification Models
Co-authored a tutorial where we modelled and validated ML pipeline design rationale using jPipe to combat "decision drift", integrating automated conformance checks into CI/CD for reproducible, requirement-traceable pipelines.
Aug 2026
ECMFA ↗ · Journal Paper
Capturing and Organizing Reusable Interaction Practices Using Justification and Feature Models
Developed a novel framework combining safety-inspired justification diagrams with feature-model variability management to make tacit best-practice knowledge explicit, composable, and reusable.
Demonstrated automated composition (assembly and refinement operators) of individual practices into coherent, auditable argumentation structures using the JPIPE and UVL/FLAMAPY toolchains.
Jun 2026
SCC-BOOK'26 ↗ · Book Chapter
Role of MLOps to validate the Requirements of AI Models using Lightweight Assurance Cases
Proposed a lightweight assurance-case approach integrating MLOps practices to support requirements validation and certification of AI systems within a book chapter.
May 2026
ICSE-NIER ↗ · Conference Paper
Model Cards for Responsible AI: Stop Carding, Start Modelling
Lead-authored a paper that evaluates current AI/ML documentation practices and introduces an actionable, verifiable alternative to narrative model cards.
Extracted and categorized 32 recurring safety patterns from provider model cards; quantified inter-provider inconsistency (only ~22% shared coverage across patterns).
Operationalized model-card content into structured assurance cases for CBRN and cybersecurity risks using evidence from GPT-OSS, Claude 3, and Gemma 3n.
Apr 2026
Education
Ph.D. in Software Engineering
McMaster University · Hamilton, ON
Topics: AI safety, Assurance Cases, MLOps
Supervised by Dr. Sébastien Mosser under the FATES-MLOps project
GPA 3.9/4.0 Supervised by Dr. Sébastien Mosser under the FATES-MLOps project
Sep 2024 – Present
Bachelor of Eng. in Software Engineering
McGill University · Montreal, QC
Sep 2020 – May 2024
Experience
Graduate Lead Teaching Assistant
McMaster University · Hamilton, ON
Sep 2025 – Present
- Lead Teaching Assistant for SFWRENG/COMPSCI 3RA3 and SFWRENG 3A04, providing academic support through office hours, tutorials, and student mentoring.
- Co-designed course materials, assignments, and assessments in collaboration with the instructional team to improve learning outcomes.
Research Assistant III
McSCert · Hamilton, ON
May 2025 – Aug 2025
-
Worked on jPipe, an open-source justification modeling language to support software maintenance activities
- Research: Designed and developed composition operators within jPipe, allowing for reusability and scalability
- ML Fairness Techniques: Implemented justification frameworks related to fairness definitions and FATES principles
- Academic Presentation: Presented the system prototype in academic settings to gather user feedback and demonstrate its role in supporting fairness-aware ML development
Software Analyst Intern
Desjardins Group · Montreal, QC
May 2023 – Aug 2023
-
Worked with the team of TFE, Electronic Fund Transfers, to modernize the current application of DRD, Direct Deposit-Withdrawal
- Quality Improvement: Successfully solved character encoding discrepancies to achieve 100% precision per 1000 files
- Query Design: Established and optimized SQL/Oracle queries to retrieve files exhibiting problematic behavior, allowing for targeted analysis and troubleshooting of identified issues
- Validation: Implemented tests to validate the effectiveness of the solutions for 4 different character encoding
Frontend Developer Intern
The Quebec Ministry of Public Security · Montreal, QC
May 2022 – Mar 2023
-
Worked on IGO2, a free open source Web Geospatial solution used by various government departments
- User Customization: Revamped the UI/UX to enable user customization, resulting in a 25% reduction in clutter
- Interactive OpenLayers Overlay: Enabled users to interact with geographical data effectively by creating 7 new features
- User-Centric Design: Prioritizing usability testing and feedback loops to align the final product with user expectations
Technical Skills
Languages
Databases
Frameworks
Cloud & DevOps
Tools
Software Projects
JavaScript · HTML/CSS · Bootstrap · jQuery · Vercel
Designed, implemented and deployed the website for the Cognitive Science Undergraduate Research Forum
Aug 2024 – Aug 2025
KnowledgeKnot — ML Engineer
Django (Python) · Angular (TypeScript) · Docker · OpenAI
Engineered a NLP model with BERT to classify & cluster info, and displaying it on an interactive web interface
Sep 2023 – Apr 2024