Postdoctoral Fellow — Machine Learning Systems École de technologie supérieure (ÉTS), Montreal, Canada PhD in Artificial Intelligence
Currently Postdoctoral Fellow at ÉTS, focusing on machine learning systems, distributed GPU training, and minimal-supervision deep learning for visual recognition tasks. PhD in Artificial Intelligence from ÉTS, contributed to the development of scalable and reproducible deep neural models for real-world applications.
At ÉTS I am also responsible for managing high-performance GPU clusters and implementing dataset delivery platforms, authentication systems, and monitoring tools to enhance scalability and reliability. Passionate about bridging the gap between machine learning research and practical deployment with a focus on advancing reproducible and interpretable AI solutions.
- Weakly and minimally supervised learning
- Test-time adaptation (TTA)
- Person re-identification (ReID)
- Domain generalization and camera bias mitigation
- Representation learning
- Distributed deep learning
- Reproducible ML systems and infrastructure
At ÉTS, I contribute both to methodological research and to the design of large-scale ML systems.
- Management of a 20+ GPU distributed training environment
- Capacity planning, performance benchmarking, and resource allocation
- Standardization of CUDA/cuDNN environments for reproducible training
- Deployment of dataset delivery platforms with versioned storage and snapshotting
- Authentication infrastructure migration (LDAP-based) with rollback-safe cutover
- Monitoring and reliability engineering via Prometheus and custom automation
Montreal, Canada LinkedIn: linkedin.com/in/smurtaza2