Case Study
CanadaCart
Product Designer & Front-End Developer • Personal Project • 3 Months
An app that helps Canadians find domestic grocery alternatives — powered by AI, built from scratch, and way outside my comfort zone.
Project Overview
This all started when I was messing around with AI tools and learning a bit of JavaScript for fun. Tariffs on Canada had just been announced, and I thought — maybe I can do something small that actually helps our country. Sounds kind of corny, but that was the spark.
The idea: a grocery list app that swaps your usual picks for Canadian-made alternatives — before you ever leave the house. Every other "buy Canadian" app I tried relied on scanning products in the store, which felt clunky in practice. I wanted something that could solve the problem before I left for the grocery store.
I'd never coded a full app before. I had designed a ton of interfaces in my career, but this was different. It was mine. I led the entire product design and front-end build, then pulled in friends to help with the database and marketing. Within 3 months, we launched an iOS app.
TL;DR
It's an AI-powered grocery list app that helps Canadians buy local — designed to be simple enough for your boomer dad to use.
The Problem
There's no easy way to know where the stuff in your cart actually comes from — especially before you even step into the store. A few apps tried to solve this, but they all relied on scanning products with a phone camera. I figured: if you're already making a list, why not have it suggest Canadian-made alternatives along the way?
The Process
I bought a whiteboard, blocked off some weekends, and invited a friend (a PM at Grammarly) to come brainstorm. We decided to begin the ideation process by setting timers to stay focused on rapid firing solution ideas. We quickly gravitated towards the grovery list idea.
I designed the app in Figma, then used a weird tag-team system between GPT and Augment to build it. GPT helped write prompts, Augment handled the front-end scaffolding, and I handled everything in between — troubleshooting layouts, writing CSS by hand, and constantly breaking and fixing the code. It was messy and extremely fun.
The tools we used:
• React Native (Expo)
• Firebase
• Algolia
• Vercel
• Figma
• A million AI coding tools
• …and Google Docs and Slack, of course.
Eventually, we brought on a developer friend to help wrangle the database and filtering logic. I stayed on as the lead designer, product owner, and front-end dev.
What It Does
CanadaCart lets users build a grocery list — and as they add items, it suggests Canadian alternatives. It uses AI to infer the product category, even with typos or brand names (e.g., "doritos" still pulls up "nacho chips"). You can scroll through alternatives, pin the one you want, and filter by "Canadian Only," "All," or "All Except USA." You can also tap into any product to see where it's made.
The entire system updates dynamically, so we're constantly improving the results in the background.
Features I'm Proud Of
• Flag-based filtering — The little country flags aren't just aesthetic. They reflect verified origin data.
• Dynamic AI search — Type "cheezies" or misspell "avocados," and it still figures out what you meant.
• Slack-integrated feedback system — If something looks off, users can hit a button, leave a note, and the app sends it (plus the screen, product, and context) directly to our Slack. The first time I saw that work, I actually laughed out loud.
• It works. That still blows my mind.
Challenges & Learning Moments
I made a ton of mistakes. I wrecked the codebase at least twice, including one 2 a.m. disaster where I accidentally reverted to the wrong GitHub repo and thought I'd lost everything. I was ready to quit, but our dev stepped in and saved the day. Crisis averted.
Learning Git, VSCode, Firebase rules, API structure, deployment pipelines — it was all new to me. But I was obsessed. I took a JS course on Udemy just to understand what I was doing better. I felt like I was learning 10x faster than usual because I had so much stake in the thing I was building.
Reflection
I've worked on a lot of products, but never this close to the guts of it all. Having skin in the game changes everything. I was designing, building, testing, breaking, fixing, learning — all at once. It was exhausting and honestly one of the most energizing creative experiences I've ever had.
This project taught me that I can build things myself — and that when I really care about something, I go hard. I also learned the value of releasing early, even if it's not perfect. Our first version was scrappy, but now we're getting real feedback, real reviews, and the app is improving every day.
What's Next
CanadaCart is live on iOS. People are using it. Reviews are trickling in. And we're already talking about how to evolve it from a "list-first" app to a "search-first" experience — something that more closely matches how people actually shop.
I'm also thinking about what's next for me. Maybe a game. Maybe another tool that solves a real problem. But I know this: I want to keep building.
