Case study · Built with AI
Making dinner easier at our house.
I built Mise en Place for my own family. We had different dietary needs, too many separate dinners, and a grocery routine that took more effort than it should.
The problem
Separate dinners, every week.
In our house, one person is vegetarian, one avoids dairy, and one follows a medically restricted diet. Put together, finding one meal that worked for everyone was rare, so separate dinners became the default and grocery shopping became a weekly chore.
I was also looking for a real problem to learn to build with AI on, and this was one I understood well enough to know whether the result was actually helping.
The approach
Recipes with swappable parts.
Each recipe has a shared base and parts you can swap, so a chicken bowl can become a tofu bowl and dairy can drop out where it needs to. The planner balances two things: reusing ingredients across the week to keep the cart small, and keeping the meals varied so the week doesn't repeat. It learns what the family actually eats.
Once the week is set, it builds a Kroger grocery cart through Kroger's API. I picked Kroger because it was the only grocer that offered free access without a developer-account approval process.
The result
One plan, one cart, one checkout.
Now we pick a plan, look over the cart, and check out. For a household that used to cook separate meals much of the week, that has made dinner a lot easier.