May 2026 - Present
Quality Engineer Co-op
CIBC - Global Technology, Data and AI - Toronto
Getting hands-on exposure to software quality and how large-scale enterprise technology is built, tested, and shipped.
00
About
Software, AI & Engineering Systems
Arnav Munipati
I am an engineering student at Western University and a Quality Engineer Co-op at CIBC.
I spend my days in software at one of Canada's largest banks and the rest of my time building things that get measured: an NBA win-probability engine tested on 87,000+ unseen game moments and an emergency-response prototype shipped in a 36-hour hackathon, with more in the works.
Greater Toronto Area
CIBC
Western Engineering
01
Experience
Experience
Roles, education, and programs that are shaping how I think and build.
Sep 2025 - Apr 2029
Bachelor of Engineering Science
Western University - Engineering | Ivey Business School - Leadership Certificate
Engineering student focused on software development, AI, and engineering systems. Building a background in leadership and business strategy through Ivey to lead technical teams and tie engineering work to business outcomes.
Jul 2024 - Aug 2025
Swim Instructor
City of Brampton - Cassie Campbell Community Centre
Delivered swim lessons and water safety training to 80+ students per session cycle while adapting feedback to skill level and keeping a zero-incident environment.
Sep 2024 - Apr 2025
Training Coordinator
FBL Canada - Brampton
Co-led four coordinators running weekly case-competition training for 60+ members, with trainees going on to place at the Canadian National Leadership Conference (CNLC), and helped raise chapter engagement by 30%.
Jul 2024
SHAD Fellow
Shad Canada
Completed Shad Canada's selective STEAM and entrepreneurship program, collaborating on real-world design challenges and presenting technical solutions to industry professionals from companies like Tesla and Ontario Power Generation.
02
Projects
Projects
Applied projects across sports analytics, AI systems, and emergency response.
AI Sports
ClutchCast AI
Predicts an NBA team's chance of winning at every moment of a live game. The platform turns raw play-by-play data from three NBA seasons into predictive signals, tests six different models against each other on games none of them have seen before, and automatically promotes the best one to make live predictions in the browser.
- Stack
- Python, pandas, scikit-learn, PyTorch, Next.js, TypeScript, Streamlit, Plotly, Flask, SocketIO, nba_api
- Evaluation
- Six models (rule-based baseline, logistic regression, random forest, gradient boosting, GRU sequence model, PyTorch neural net) ranked by Brier score, log-loss, ROC-AUC, and accuracy on 87,694 game states across 180 games held out from training, from a 900-game three-season sample
- Result
- The neural network won: 73.1% accuracy on games it had never seen, with the strongest probability calibration of the six (0.174 Brier score, 0.815 ROC-AUC)
NVIDIA Spark Hackathon
Delation
Emergency-response prototype built at the NVIDIA Spark Hackathon in Toronto. Incident reports arrive as voice, video, and location; the system layers them with city data, flood maps, and live weather to flag which neighbourhoods are most at risk, all running locally with no cloud required.
- Stack
- Python, FastAPI, Uvicorn, Pydantic, Faster-Whisper, Kokoro ONNX, WebSockets, Next.js, TypeScript, Mapbox GL JS
- Role
- One of five engineers on the team. Built the FastAPI gateway, the system's front door: it receives every incident report, converts speech to text and back, streams live status updates, and merges Toronto Open Data, floodplain maps, and Environment Canada weather into a risk score for each neighbourhood.
- Outcome
- Shipped a fully working prototype inside the 36-hour hackathon window, built to analyze an incoming incident in under a tenth of a second on the team's on-site NVIDIA hardware.
03
Contact
Contact
Open to conversations about software, AI, and engineering systems.
Open to software, AI, and engineering internships/co-ops.
Best way to reach me is email.