
Overview
Chefy Menu is a web app that lets restaurants generate digital menus instantly from images using AI.
The goal was to remove the friction of manual menu creation and make onboarding as fast as possible.
The problem
Restaurants typically rely on static PDFs or manually entered menus, which are slow to update and painful to manage. Existing tools often require too much setup, leading to low adoption—especially for small businesses.

My role
Implemented the system end-to-end, from initial architecture to production deployment, including AI integration, data modeling, and infrastructure.
What I built
AI pipeline: image → structured menu data (sections, items, pricing), with schema enforcement, parsing, and fallback handling for unreliable outputs
Domain model: Menu → Sections → Items, designed to support both generated and manually edited content, with future extensibility in mind
Editing experience: real-time UI with optimistic updates and deterministic reordering, balancing responsiveness with data integrity
Authentication & onboarding: email and Google auth, with a streamlined flow that minimizes friction and avoids redundant uploads
Billing system: Stripe integration covering plans, trials, webhooks, and subscription lifecycle management
Infrastructure: React + Remix application, Prisma + SQLite with LiteFS replication, deployed on Fly.io with clear separation of read/write paths

Reflections
Chefy Menu reflects my approach to building products: start from first principles, keep the system simple, and focus on delivering real value quickly. From AI integration to infrastructure, every decision was made to balance reliability, performance, and user experience.
2 likes · 12 views · Apr 6, 2026
Built an AI-powered product end-to-end, from architecture and data modeling to production infrastructure.