posted an update

Day 1-2: From idea to full-stack app in 48 hours

When I saw the H0 Hackathon announced, I knew immediately what I wanted to build. As someone who researched AI-driven food intelligence for my MSc dissertation — building ingredient embeddings, graph neural networks, and novel recipe generation — I've always believed food AI could do more than just recommend popular recipes. It should work for everyone, including the 250 million people worldwide living with food allergies.

Fable is an allergen-aware recipe discovery app powered by Epicure — the largest multilingual food embedding model ever built (4.1M recipes, 7 languages, 1,790 ingredients compressed into 2MB of vector space).


Allergen picker

Allergen picker

EU Big 14 allergen filtering with a hand-curated truth table across all 1,790 ingredients. Edge cases handled correctly — oat milk is safe for dairy allergy, almond milk is not safe for tree nut allergy.


AI recipe generation

Recipe generation

Anthropic Claude generates novel restaurant-quality recipes based on your ingredients, allergen profile, meal type, and cook time. Every recipe is unique — powered by Epicure's flavour embeddings finding what works together.


Safe Foods Mode

Safe Foods Mode

Built specifically for MCAS and severely restricted diets. Define exactly what you can eat and Fable generates recipes strictly within that list. "Liquid of choice" placeholders for users who can't even assume water is safe. Nobody else is building recipe tools for this community.


Kitchen management

Kitchen management

Track ingredients by area (fridge, freezer, cupboard, pantry), set use-by or bought dates with automatic shelf life calculation, add quantities and subtypes. "Use my kitchen only" toggle constrains recipes strictly to what you have.


Tech stack

Next.js 16 · TypeScript · Tailwind CSS · AWS DynamoDB · Anthropic Claude · Epicure Core embeddings · Vercel

Try it

Live app · GitHub

Log in or sign up for Devpost to join the conversation.