Generational Gravy

Thumbnail
Home Page
Recipes
Problem Statement
Solution
Target Audience
Ethics
User Flow + Features

Inspiration
We lived it—six months abroad, then moving out of Montreal for work. Craving home cooking hit hard. That gap became Generational Gravy.

What it does
Instantly recreates your family's signature dishes anytime, anywhere—no recipe required.

How we built it
Claude for smart prompting, Twelve Labs for video understanding, OpenAI for recipe extraction, and TypeScript for a rock-solid frontend.

Challenges we ran into
Computer vision inconsistency (identifying "Mom's pinch of salt" is hard), API rate limits, and matching vague family recipe descriptions to real ingredients.

Accomplishments we're proud of
Seamless multi-API integration. Got CV to recognize 80%+ of dish variations. Shipped a working MVP in 36 hours.

What we learned
Same as always: scope tighter than you think, test CV with real user photos early, and never trust ingredient detection on the first pass.

What's next
Train our vision model on 10x more home-cooked dish photos. Add voice recording for family storytelling. Launch beta to college students.