Fresh Local Food Marketplace 🥕

BuyFresh.Food - Where fresh means local.

What Inspired Me

I was buying eggs from my neighbor's daughter's friend who keeps chickens, and they were struggling to keep up with demand. That got me thinking - if I knew more people selling food locally, I'd probably buy from them too.

Farmers markets are wonderful, but they're only open 6 hours a week and usually on weekends when I'm working on home projects or trying to rest. I started noticing my friends driving 30-40 minutes across town just to get whole milk from one specialty farmer, while other friends who picked up gardening and baking during COVID always had tons of leftovers.

It hit me that there's this weird disconnect - people want local food but can't find it easily, and people making food don't have a simple way to sell it. As a product manager, I realized I could probably build something to fix this.

What I Learned

Building this project taught me I could actually make a real app using AI tools. I couldn't believe I was able to build something with a database, API calls, user accounts - the whole thing.

The biggest learning was figuring out how to work with these AI tools effectively. Claude helps me write better prompts, Subframe lets me design like I'm in Figma, and Bolt turns it all into working code. But getting them to do exactly what I wanted took practice - I had to learn to be really specific and work one piece at a time.

What really excited me was realizing these skills make me better at my day job as a product manager. Now when I talk to engineers and designers, I actually understand what they're dealing with - design tokens, state management, all that stuff I used to just nod along with.

How I Built It

I worked on this nights and weekends using a workflow I developed:

Plan in Claude - I'd describe what I wanted to build and have Claude help me think through the user experience and write prompts for the design tool
Design in Subframe - Take that prompt and build the actual interface, iterating on layouts and components just like I would in Figma
Code with Bolt - Export the design, have Claude create a detailed prompt for Bolt, then implement everything in their React environment
Test and deploy - Test the features right in Bolt, then push to Netlify to see it live

The whole process let me focus on solving the actual problem instead of wrestling with code syntax.

Tech I used:

React + TypeScript + Vite
Subframe for design system
Supabase for the database
Netlify for hosting
Google Places API for locations

What I Built

The app has all the features you'd expect from a modern marketplace:

Splash pages with CTAs - Clear landing pages that guide users to key actions
Search - Find local producers and products
Add Location - Set your area to discover nearby food sources
Waitlist - Join early access for your region
Shop using a map - Visual discovery of producers in your area
Product Page - Details about products and who's selling them
Seller Page - Producer profiles with all their offerings
Shopping Cart - Add items and coordinate pickup
Schedule pickup - Coordinate convenient pickup times
Works on all devices - Responsive design for mobile and desktop

Challenges I Faced

The hardest part was learning to communicate with the AI tools. When I'd ask Bolt to change something, it would sometimes break other parts or not understand exactly what I meant. I learned to work incrementally - change one small thing at a time and be super specific about what to keep and what to modify.

I really struggled with syntax errors and TypeScript state management logic. The bugs took forever to fix at times - lots of discussions between me and all of my AI tools. LOL. Getting the location features working was also tricky. I wanted different experiences for people in different areas, and making sure the Google Places API worked smoothly took some back-and-forth.

Since I'm familiar with databases and SQL, the Supabase setup wasn't too bad, but connecting everything properly and making sure the frontend talked to the backend correctly definitely had its moments.

Why This Matters

This project showed me that regular people can build real solutions to real problems. I'm not a developer, but I was able to create something that could actually help my community connect around local food.

The bigger picture is that tools like this are changing who gets to build things. As a product manager, being able to prototype and test ideas myself instead of always needing engineering resources is pretty powerful.

We're starting in Northwest Arkansas to test some key assumptions: Are people willing to schedule pickups rather than full delivery? Will farmers actually sign up for yet another tech solution? Can we attract cottage food producers, hobbyists, homesteaders, and gardeners along with actual farmers - maybe even more easily? What are the key products and assortment that keep people coming back? Once we figure out what actually works, then we'll expand from there.

URL: https://buyfresh.food/
Demo: https://youtu.be/nY3Tq63nHWo?si=aN87sa43QnXGl5nK Built with: Claude + Subframe + Bolt.new

Hackathon Entries: