Sagan Labs AI Logo
SkanChef — Your fridge, your chef.
SkanChef logo

SkanChef · Consumer Food-Tech / AI Kitchen Assistant

SkanChef — An AI Recipe App That Cooks From What You Already Have

Your fridge, your chef.

What it does

AI ingredient scanning (fridge or pantry photo to items)
Grocery-receipt OCR
AI recipe generation matched to on-hand ingredients
Weekly meal plans and smart shopping lists
Cook-Along guided mode with timers
Cross-platform: iOS, Android, and web (PWA)

Built with

React 19
TypeScript
Vite
Tailwind CSS
Firebase (Auth / Firestore / Cloud Functions)
Google Gemini (vision)
Capacitor (iOS / Android)

SkanChef is an AI recipe app built by Sagan Labs AI that turns a photo of your fridge into dinner. Point your phone at what you already own, and computer vision reads the ingredients, then generates recipes you can actually cook tonight — no separate shopping trip, no staring at a full fridge with nothing to make. It is a live consumer product (skanchef.com) and a working demonstration of the custom AI app development we do: production vision AI, wrapped in an app people use every day.

The Problem

The hardest part of cooking at home is not the cooking — it is the decision. People open a stocked fridge, fail to see a meal in it, order takeout, and let good ingredients spoil. The tools meant to help usually work backwards: pick a recipe first, then go buy things you do not have. That adds cost and waste to a problem that started with food already in the house.

Our Approach

SkanChef starts from what you own, not from a recipe you have to shop for. Snap a photo of the fridge, pantry, or even a grocery receipt; Gemini vision detects the ingredients with confidence scores, and you confirm or edit the list. From there the app generates recipes matched to those ingredients and your dietary filters, then carries the result forward into meal plans, an auto-built shopping list for anything missing, and a step-by-step Cook-Along mode with timers.

All of the AI runs behind a server-side proxy, so the vision and generation calls happen safely in the cloud rather than on the device — the same architecture we use when we ship AI features for clients.

Architecture Highlights

  • Vision-first input. A single photo of your ingredients drives detection with confidence scoring and a human review step, so the recipe is grounded in what you actually have.
  • Server-proxied AI. Gemini calls for detection, receipt OCR, and recipe generation run through Cloud Functions, keeping keys and model access off the client entirely.
  • One codebase, three surfaces. A React and Vite PWA packaged with Capacitor ships to iOS, Android, and the web from a single product.
  • Multilingual and offline-aware. English, Spanish, and Brazilian Portuguese, with a responsive, install-to-home-screen experience.

Outcomes

SkanChef's goal is to remove the "what's for dinner?" decision cost and turn a full fridge into a made meal — reducing food waste as a side effect.

  • Turns a single photo of your fridge into cookable recipes matched to what you own.
  • Ships as one product across iOS, Android, and web from a shared codebase.
  • Live in production at skanchef.com, in English, Spanish, and Brazilian Portuguese.

Ready to build an AI app like SkanChef?

SkanChef is proof that Sagan Labs AI ships real consumer vision AI — not a demo, a product people cook with. If you have an idea that turns a camera into an experience, we build the same way: production model access proxied safely, one codebase across mobile and web, and a real product on the other side. See how we build AI products, talk to us about your app, or browse the rest of our work.

Want AI like SkanChef built around your operation?

Sagan Labs AI designs and ships production AI systems — computer vision, document AI, agents, and automation — tailored to how your business actually runs.

Request a consultation