Canada SOP for Nigerians — Statement of Purpose for Master’s & PhD

Statement of Purpose — MSc Computer Science, University of Toronto

In 2021, I built a payment reconciliation system at Paystack that processes ₦400 million monthly across 12,000 Nigerian merchants. What started as a 3-month internship became a 4-year journey through every layer of distributed systems engineering. But somewhere between debugging race conditions in Kafka and reading the original Chubby paper, I realized I wanted to do this work at a deeper level. That is why I am applying to Toronto, with specific interest in Professor Yashar Ganjali’s research on programmable networks.

My fascination with distributed systems began in my third year at the University of Lagos, when I built a Yoruba speech-to-text model. The model worked, but training took 14 hours on a single GPU. When I distributed training across 4 machines using PyTorch DDP, time dropped to 90 minutes. That gap — between sequential and parallel computation — has stayed with me.

At Paystack, I worked on a real-world version of this question. Our reconciliation service was failing 8% of the time due to race conditions. I introduced a Saga pattern and reduced failure rate to 0.4%. But I learned what production distributed systems actually demand: not just correctness, but observability, recoverability, graceful degradation.

I want to extend this work at Toronto specifically because Professor Ganjali’s 2019 paper on stateful network functions (NSDI 2019) addresses the problem I face daily: how to maintain consistent state across geographically distributed services without sacrificing latency. I would be honored to contribute to his current project on AI-driven network optimization.

After my degree, I plan to either pursue a PhD or return to industry research labs (Google Research, Microsoft Research) where the theory-practice gap is smallest. I’m committed to bringing what I learn back to the African tech ecosystem.