Research / engineering portfolio
I develop machine-learning systems that are interpretable, theoretically grounded, and efficient enough for real hardware.
Now
Incoming MMath
Waterloo / Sept 2026
Focus
Explainable AI
Theory + hardware
Output
Rail PdM pipeline
Preprint / 2025
About
Research premise
If we cannot explain why a model works, we cannot trust it when it matters most.
I graduated from the University of Western Ontario with a B.E.Sc. in Software Engineering. In September 2026, I begin an MMath in Computer Science at the University of Waterloo, supervised by Dr. Robin Cohen and Dr. Lukasz Golab.
My work moves between physical constraints and mathematical ones: spiking neural networks on FPGAs, on-device 3D reconstruction, railway predictive maintenance, and interpretable learned decisions.
Toolkit
Python/C/C++/C#/VHDL/TypeScript/PyTorch/TensorFlow/React/Next.js/Node.js/FPGA/MATLAB
Research directions
Making neural-network decisions transparent and trustworthy through feature attribution, concept-based explanations, and mechanistic views of learned representations.
Feature attribution / Concept explanations / Trust in AI
Studying loss landscape geometry, training dynamics, and implicit regularization to understand why deep networks generalize.
Loss landscapes / Implicit bias / Training dynamics
Bridging models and physical constraints through spiking neural networks on FPGAs and efficient inference for edge devices.
Spiking networks / FPGA acceleration / Edge inference
Publication
PUB_001
Preprint
2025
Rail PdM
Xristopher Aliferis, Farzan Heidari, Tangjian Wei, Yili Tang
Preprint / 2025
Experience
Education
2021 - 2022
Bachelor of Science
Computer Engineering / College of Engineering
2022 - 2026
Bachelor of Engineering Science with Distinction
Software Engineering / Electrical and Computer Engineering
Starting Sept. 2026
Master of Mathematics (MMath)
Computer Science / David R. Cheriton School of Computer Science
Supervised by Dr. Robin Cohen and Dr. Lukasz Golab.
Honors
Featured result / 1st place
INFORMS Railway Applications Section / 2025
First place in the 2025 INFORMS Railway Applications Section Problem Solving Competition for developing a predictive maintenance framework using machine learning on railway detector data.
Research funding / NSERC USRA
Competitive national research award supporting undergraduate work in neuromorphic computing and 3D head reconstruction under Dr. Roy Eagleson.
Competitive national research award supporting undergraduate work in machine learning and predictive maintenance under Dr. Yili Tang at the MoTech Group.
Additional academic recognition
Projects
Self-directed collaborative manuscript testing whether LoRA fine-tuning can help a 1.5B LLM approach a 72B model on multi-step algorithmic puzzles, using solver-generated data and structural metrics.
Cal Hacks 12.0
MLH - Best .Tech Domain Name
A gamified productivity platform built in Godot that transforms task completion into world-building. Integrates AI assistants to break down complex tasks and motivate users.
UC Berkeley AI Hackathon 2025
A social networking platform where personalized AI agents act as digital twins, initiating conversations and surfacing meaningful connections.
