Dinner was at six. It's almost seven.
It's 6:52 on a Tuesday. Dinner was supposed to be at six. The chicken is finally done, but the rice went cold twenty minutes ago and the vegetables are sitting somewhere in between. You did everything the recipe said. Nobody's eating a hot plate tonight anyway.
That gap — between the plan and what actually happened at the stove — is the whole reason I'm building SousSmart.
Recipe timing is a guess wearing a fact's clothes
A recipe hands you numbers and dares you to trust them. Bake 45 minutes. Rest 20. Simmer until reduced. They read like facts, but they're averages from a test kitchen that isn't yours. Your oven runs cool. You opened the door three times. The bird was a pound heavier. You started late because the day happened.
Every one of those numbers is an estimate, and the moment your kitchen touches it, it's wrong by some amount you're now expected to do in your head — while your hands are covered in raw chicken.
Recipe apps store. Timers count. Neither one cooks.
Look at the tools we have. A recipe manager is a nice place to keep the recipe, and then it hands you a wall of text and walks away. A multi-dish timer counts down the fixed numbers you punched in, and hopes. Both of them make the same quiet assumption: that the plan you started with will survive contact with the kitchen.
It never does. And the second one thing slips, you're on your own again, recomputing when to start the potatoes so they're not cold, holding the whole meal together in your head. That's the job nobody built a tool for. That's the job I want to build.
Throw out the timing. Keep dinner on time.
SousSmart doesn't throw out the recipe. It throws out the fiction that the recipe's timing is fixed. You tell it one thing — when you want to eat — and it builds the plan backward from that moment, around your dishes and your kitchen, not a stranger's.
Then it stays with you. Finish a step early and the plan moves up. Start late, hit a slow boil, pull the turkey when it's actually done instead of when a chart said so, and everything downstream re-times itself so dinner still lands together. You never recompute in your head, because the app is holding the plan for you. The timeline is the product. The recipe is just where it starts.
I keep complicated things on schedule for a living
I've spent my career running operations where a lot of moving parts have to land at the same moment, and reality refuses to cooperate. What I learned is that the plan is never the hard part. Holding the plan when everything shifts under it — that's the whole job. A holiday dinner is that exact problem, shrunk down to one kitchen and one very specific six o'clock.
It's also a problem you can finally solve well on a phone. The pieces to do this — on-device, offline, fast enough to keep up with a real cook — only recently fell into place. So the timing, for once, is right.
I'm building it in the open
This is the build log, and I'm going to write it honestly. Every build, I cook a real meal through the app and tell you what broke — the first thing a real kitchen broke was a three-hour roast that trapped me in my own screen, and that story is already up here. You'll get the decisions, the dead ends, and the reasons behind them.
The one thing I'll hold back is the part that's hard to copy — how the re-planning actually works under the hood. You'll see the what and the why, not the how. Everything else is on the table.
You told me dinner's at six. The whole app is just trying to keep that one promise