ML Research Assistant at 
GeoTrans Labs
I'm a 3rd Year Computer Engineering Student. I live and breathe programming.
GeoTrans Labs. Previously, I built production software as a Software Engineer Intern at 
VSF Holdings, completed my SWE (AI/ML-focused) Residency at 
Headstarter, and wrapped up my role as an AI Student Ambassador at 
Intel Corporation. I’m interested in full-stack or backend/systems engineering roles where I can build and ship at scale. I’m eager to break into banks or big tech companies serving millions - if not billions - of users. I’m currently searching for an 8-month co-op starting Winter 2026, and I’m open to longer terms when the fit is right.
Python, 
TypeScript, 
JavaScript, 
C/
C++, and shells things together with 
Bash. I craft the client-side with 
HTML, 
CSS, and ship polished interfaces in React, Next.js, and rapidly iterate with Tailwind. On the backend, I spin up services in Node.js, Express.js, and fabricate APIs with 
FastAPI. Data lives in PostgreSQL, MongoDB, Supabase, and Firebase, with plenty of monitoring and testing sprinkled in.
. GeoTrans Labs, University of AlbertaSep 2025 - Present
Building GeoTrans Lab’s road-surface and tire-friction monitoring platform end-to-end - ingesting traffic and weather data, running Swin Transformer inference on batched imagery, and streaming results into mapping dashboards so our stakeholders can spot slippage risks in real time. I tuned the ONNX + FastAPI pipeline with GPU acceleration to drop latency to under five seconds for 144 images and layered in Gemini-based preprocessing so we can skip frames with untrackable conditions, pushing inference accuracy to 94% across five surface types. Everything ships through Dockerized deployments on Cloud Run backed by GitHub Actions, which cut our setup time by almost half.
VSF HoldingsJun 2025 - Aug 2025
Partnered closely with VSF’s executives to design and ship a full-stack investment analytics platform on Vercel using Next.js, TypeScript, and Node/FastAPI services. I owned the UX for investor dashboards, session analytics, and CRM/KYC onboarding flows, which ultimately tripled average visit duration and slashed manual onboarding work by 60% thanks to automated ETL pipelines in Python and SQL. Weekly Agile sprints ran like clockwork - design reviews, Figma prototyping, and stakeholder demos every Friday kept us shipping on time for every release.
IntelMay 2025 - Aug 2025
As an Intel AI Student Ambassador I led hands-on workshops that walked 300+ learners through the AI Analytics Toolkit, Modin, and XGBoost optimizations - everything from profiling data pipelines to squeezing extra accuracy out of boosted trees on Intel hardware. Outside the classroom I build demo projects on Developer Cloud, publish performance case studies, and share ready-to-run notebooks on Hugging Face so the community can replicate the optimizations.
HeadstarterMay 2025 - Aug 2025
As a Software Engineering Resident at Headstarter, I’m building 14+ full-stack and AI/ML projects focused on deploying real-world tools. My work spans LLM chaining, hyperparameter tuning, and fine-tuning over 10+ models, all with production-ready pipelines. I’ve contributed to both Python-based backends and TypeScript frontends, and have been mentored by engineers from Google, Tesla, and Two Sigma.
Computer Engineering Club, University of AlbertaSep 2024 – Present
As VP External I run industry outreach for Alberta’s busiest student tech organization - lining up sponsors for Aether, HackED, and HackED Beta, co-hosting multi-club workshops and hackathons, and building the events site that keeps 2,500+ students in the loop. Before stepping into the VP role I served as a Junior Executive for the 2024–25 term, representing second-year Computer Engineering students, mentoring incoming members, and spearheading the initial sponsor pipeline that we still rely on today.
AlbertaSATSep 2024 – Dec 2024
Contributed to AlbertaSAT’s 2026 CubeSat launch as a software member on an Agile squad, pairing with electrical and mechanical teams to vet subsystem code long before we risked hardware testing. I built Rust-based simulators for more than two dozen components, helping uncover four critical integration issues and keeping the mission timeline intact.
Analyzing winter road grip conditions as a GeoTrans Lab undergraduate research assistant.
AI-powered ergonomic wellness system combining computer vision and EMG sensors for real-time posture, eye strain, and wrist monitoring. Built for Nathacks 2025.
Real-time posture monitoring and analysis using AI-powered pose detection, built during Hack The North 2025.
AI-powered property recommendation system for real estate appraisals
Classifying brain tumors in MRI scans using advanced deep learning, with a Streamlit web app and Xception transfer learning.
An intelligent system that powers automated customer service interactions through AI-driven conversations
Multimodal video assistant powered by Gemini - ever wish your video could answer back? Now it can
A Python-based tool that analyzes Reddit posts and comments to gauge stock sentiment, correlating social sentiment with market data for trend insights.
End-to-end cointegration-based pairs trading engine with clustering, OLS modeling, ADF testing, rolling z-score signals, and a full portfolio backtester.
AI-driven medical form automation system that extracts data from referral packages and fills Prior Authorization PDFs in under 2 minutes
EMG-based gesture control interface using the Bioamp EXG Pill and Arduino. Maps forearm muscle contractions to keyboard actions with real-time signal processing and peak detection algorithms
Your Python-based AI sidekick: Gemini does the heavy lifting, then pushes polished summaries into Docs
