I work on the co-design of
machine learning models and
the systems that run them.
I'm a fourth-year Computer Science (Specialist) undergraduate at the University of Toronto, focused on the intersection of computer systems and machine learning. My research interests center on the interaction between ML models and their underlying systems and hardware — improving the models and the kernels, compilers, and systems infrastructure that run them as a single co-design problem.
Currently, I'm a Research Assistant at the embARC Lab with Prof. Nandita Vijaykumar, developing optimization frameworks for GPU kernels for cryptographic workloads. I am applying to PhD programs to continue working on systems for efficient, reliable, and secure machine learning.
Experience
Academia
May 2026 – Present
Research Assistant
embARC Lab · University of Toronto · Toronto, ON
Cryptographic workloads need kernels as specialized as the models they serve.
Building optimization frameworks for GPU kernels used in cryptographic workloads. I write kernels for finite field arithmetic (e.g., multi-scalar multiplication) and develop profiling infrastructure and theoretical models for hardware-independent optimization across architectures.
January 2026 – May 2026
Research Assistant
SITH Lab · University of Toronto · Toronto, ON
Modern GPU memory stacks have security properties we do not fully understand.
Developed custom CUDA/HIP kernels to parallelize attack execution, reverse-engineered memory and bank layouts via timing analysis, leading to a 23,500x increase in Rowhammer bit-flips. This work was accepted to ACM CCS 2026.
August 2025 – May 2026
Research Assistant
Machine Intelligence Lab · University of Toronto · Toronto, ON
Memory mechanisms in brains can inspire more efficient and robust models.
Engineered Transformer layers in PyTorch with customized self-attention, scaling across a 50M-parameter architecture, and built optimized data loaders and multi-threaded pipelines to preprocess 15,000+ Wikipedia articles. Findings are under review at NeurIPS.
May 2024 – April 2025
Research Assistant
Vector Institute, UHN · Toronto, ON
Building synthetic data generation pipelines to address data scarcity in healthcare.
Developed image segmentation software for image annotations and preprocessing, experimented with inpainting models to remove spurious correlations in X-rays, and built a classifier for pneumothorax triaging using PyTorch on UHN's GPU cluster.
Industry
June 2025 – December 2025
Software Engineering Intern
Mezzi · San Jose, CA
Built scalable Go backend APIs and a Redis/Firestore caching layer to serve financial data aggregation.
June 2025 – August 2025
Software Engineering Intern
Tigera · San Jose, CA (Part-time)
Shipped distributed network validation microservices and a GCP RAG backend for network-policy operations.
Publications
- 2026
GPUThor: Amplifying Rowhammer Attacks via Non-Uniform Patterns to Exploit ECC-Protected GPUs
C. S. Lin, J. Qu, A. Rajeev, G. Saileshwar
Proc. ACM Conference on Computer and Communications Security (CCS)
Projects
Education & Awards
Education
Sept. 2023 – May 2027
BS, Computer Science, focus in Computer Systems and AI
University of Toronto
GPA: 3.99/4.0 (Dean’s List)
Teaching Assistant, Systems Programming (CSC209) — Winter 2026
Awards & Honors
- 2026
Department of Computer Science Award
$12,000
- 2024
University of Toronto Excellence Award
$7,500
Dean’s List
University of Toronto
Relevant Coursework
- CSC413Neural Networks & Deep Learning
- CSC469Operating Systems Design
- CSC488Compilers
- CSC311Introduction to Machine Learning
- CSC368Computer Architecture
- CSC369Operating Systems
- CSC209Systems Programming
- CSC258Computer Organization
- CSC373Algorithm Design
- CSC263Data Structures & Algorithms
- CSC343Introduction to Databases
- MAT223Linear Algebra
Personal Interests
Contact
I'm always glad to discuss systems for ML, hardware-software co-design, or graduate research opportunities. The fastest way to reach me is email.
Aditya Rajeev · Toronto, Canada · Last updated 2026