โ† Back to all posts

CookShare: From Their Kitchen, to Yours

๐Ÿฒ

CookShare is a small social app I've been building for home cooks. The tagline โ€” "From their kitchen, to yours." โ€” is the entire pitch. Most food apps optimize for restaurants. This one optimizes for the auntie down the street whose biryani is better than anything you'll order on a delivery app.

The product, in one paragraph

A home cook posts what they're making this week. Neighbours subscribe, comment, request, and order in small batches. The cook isn't running a restaurant โ€” they're feeding a few extra portions to people who already trust them. The platform's job is to make that trust visible, not to manufacture it.

The stack

  • Laravel 12 on the backend โ€” fast, opinionated, and I don't have to invent auth or queues.
  • Flutter 3.38 for the mobile app โ€” one codebase, both stores.
  • MySQL + Redis for the boring, durable parts.
  • S3-compatible object storage for the food photos, which are the entire content surface.

Design decisions I'd defend

  1. Cooks are the first-class citizen. The home screen is people, not dishes. You follow a cook the way you follow a creator on Instagram, not a menu.
  2. No ratings. Stars destroy small communities. Comments and re-orders carry more honest signal.
  3. Geofenced by default. A cook 12 km away is not a useful result. Most of the magic happens within a 2 km radius.
  4. Scheduling, not "available now". Home cooks aren't a 24/7 supply. The product respects that.

The hard parts

The technical work is easy. The hard part is the trust layer โ€” verifying cooks without making it feel like a bureaucratic onboarding. I'm currently testing a "trusted neighbour" flow where existing customers vouch for new cooks, the same way home-cook businesses already work in real life. That's the kind of thing I'd never have figured out from a whiteboard.

Every food app eventually becomes a logistics company. CookShare is a deliberate bet that there's space for one that stays a community.

What's next

A small Dhaka pilot first โ€” three or four neighbourhoods, fifty cooks, very high-touch onboarding. If the retention curve looks like a community and not a marketplace, I'll know I've built the right thing.