AI assistant · N5 Complete system
Prep Procedure Builder for Restaurants
Takes the way a cook makes each preparation out of their head and writes it down step by step, with time, control signal, batch size and shelf life, so the dish stops depending on who is in that day.
N5Complete system
Several chained pieces you install and operate over time, with their review cadence. It changes how the business runs, not one decision.
◆ This assistant 110 of the 164 published assistants are at this levelIs this for you?
What it does and what you get
- Almost every kitchen has one preparation that only comes out right when one particular person makes it, and almost no owner calls it what it is: a dependency, paid for the day that person falls ill or quits.
- This assistant turns "you just know when it's ready" into steps with a time and a control signal, states how many portions the batch is written for, measures process waste and closes with the stretch nobody documents: chilling, labelling and shelf life.
What you end up with
A step-by-step procedure sheet with time and control signal; the batch table with portions and process waste; the night-before prep checklist; the chilling and shelf-life table; and a two-week handover plan with owner, date and indicator.
EXAMPLE This is what you getAn example built with Sazón de Origen, the house's fictional restaurant.Not your data: your result is built with YOUR restaurant's.See the full example →Sample result
The owner asked this: «The three-meat sancocho has been made by one person for six years. Last week she was off sick for three days, on Sunday it came out watery and we took four back. I need the procedure written down so anyone on the team can run it, and I want to know which stretch does not fit into the shift.»
Eleven steps for a batch of 21 portions. The three in bold are the ones that used to live inside a sentence nobody else could execute.
| Step | Action | Time | Control signal | Equipment | Who |
|---|---|---|---|---|---|
| 1 | Take the three meats out of the walk-in and temper | 20 min | The thick piece goes from 3 °C to 8 °C at the core | Cold table | Pot cook |
| 2 | Bone and cut into 40 g pieces | 35 min | 8.00 kg purchased give 5.52 kg servable | Board and knife | Pot cook |
| 3 | Blanch and discard the first water | 12 min | The water runs cloudy with grey foam | 40 l pot | Pot cook |
| 4 | Cover with 9 l of clean water and bring to the boil | 18 min | A rolling boil, not edge bubbling | 40 l pot | Pot cook |
| 5 | Cook the meats | 95 min | The knife goes in without resistance on the thickest piece; it used to say "until it softens" | 40 l pot | Pot cook |
| 6 | Add cassava and plantain in 60 g pieces | 25 min | The cassava opens under the spatula | 40 l pot | Pot cook |
| 7 | Add corn in 4 cm rounds | 12 min | The kernel gives on the bite, no starch in the centre | 40 l pot | Pot cook |
| 8 | Reduce the broth | 15 min | 9 l become 6.8 l; the broth coats the back of the ladle and holds a finger line. It used to say "when it thickens up" | 40 l pot | Pot cook |
| 9 | Adjust salt and coriander | 4 min | 18 g of chopped coriander per 6 l; it used to say "a good handful" | 1 g scale | Pot cook |
| 10 | Rest off the heat | 10 min | Fat rises and is lifted off with a ladle | 40 l pot | Pot cook |
| 11 | Plate to order | 3 min | 260 g of meat, 300 ml of broth, 150 g of vegetables | 150 ml ladle | Pass |
Step 8 is the one that sent four plates back on Sunday. "Thickens up" is not a signal: it is a memory. With the usual cook the broth came down from 9 l to 6.8 l because she knew how long to leave it uncovered; with another hand the sancocho went out with 8.1 l in the pot and the same meat weight — more liquid, same protein, watery plate. The fix is not a temperature or a timer: it is a level mark painted inside the pot at 6.8 l.
| Preparation | Base batch | Portions | Process waste | Cost per portion |
|---|---|---|---|---|
| Sancocho de tres carnes | 8.00 kg of meat + 9 l of water | 21 | 31% boning · 24% broth evaporation | $9,143 |
The 24% of broth that evaporates is not a loss: it is step 8, which is why it lives inside the procedure and not on a waste list. The 31% from boning is a cost, and it already sits in the dish's cost sheet.
Doubling the batch does not double the times, and that is the trap.
| Batch of 21 portions | Batch of 42 portions | |
|---|---|---|
| Boning (step 2) | 35 min | 70 min |
| Cooking the meats (step 5) | 95 min | 120 min |
| Reduction (step 8) | 15 min | 28 min |
| Long stretch total | 145 min | 218 min |
| Does it fit the 240 min window? | Yes, with 95 min to spare | Yes, with 22 min — and no room for anything else |
SUPUESTO: the double-batch times do not come from a scaling rule of the trade, which does not exist at that precision: they come from two timed runs in this kitchen's 40 l pot. If Sazón de Origen buys an 80 l pot, step 5 gets measured again before being written, because a bigger mass takes longer to reach the boil and less to hold it. And if the double batch runs in two 40 l pots at once, the times are the ones in the first column but two hands are needed, which is a staffing decision —24 people in all, 18 in operations— and not a cooking one.
Chilling, packaging, labelling and shelf life. The part of the procedure that never shows on the plate and the only one that can close a restaurant.
| Preparation | Packaging | Label | Shelf life | What voids it |
|---|---|---|---|---|
| Strained base broth | 6 l tray, lidded | Date + chilling time + initials | 72 hours at 3 °C | It did not go from 65 °C to 10 °C in 120 min |
| Drained cooked meats | 2 kg bag, laid flat | Date + time + lot | 48 hours at 3 °C | It was reheated twice |
| Cooked cassava and plantain | 3 l tray | Date + time | 24 hours at 3 °C | It changed colour or released water |
| Plated sancocho | Not stored | — | 0 hours | Always plated to order |
The 120 min rule is not decoration: it is what separates a 6 l tray of broth chilled properly from one that spent eight hours in the danger zone because it went into the walk-in hot and covered. And the label carries the time, not just the date, because "Monday" does not say whether it is 12 or 36 hours.
Illustrative example generated with a fictional restaurant. Not a promise of results, not a client case.
This is 3 of 4 parts. Behind the wall is **the handover**: the night-before checklist with cut-off time and signature, the 15 day plan setting out what the one who knows and the one who learns each do per week, **the timed test with the cook who did not know** —210 min against 155 and 9 questions that rewrote three steps—, the closing indicator with its tolerance, and the list of the other six preparations that depend on one person today.
See the full example — free accountWorks with these AIs
How to use it
Which of your data it uses
This assistant works with 8 of the 23 fields in «My Restaurant», among them Cuisine or concept, Number of locations and Best-selling dishes. It does not ask out of curiosity: these are what make the answer speak about YOUR scale instead of an industry average — the same calculation on a business of another size returns a number you cannot decide with. You fill them once and they apply across the library, so the second assistant you open already starts with them in place. Whatever you leave blank, the assistant asks for one thing at a time instead of inventing it, and says so before answering, so you know what it is working from.
Always up to date
Every assistant carries a visible version and date. When an AI changes how it works, the assistant's version goes up and you see what changed.
v1.0.02026-08-26The procedure is born. The recipe says what goes in and how much; this piece says how it is made and in what order, which is what walks out the door when the cook who knew it quits.More assistants in this category
Compares the same recipe costed on two dates, splits every difference into price effect and quantity effect, a…
N5 · Complete systemHolds every spec sheet against real service with four tests —the weighed plate, the photo, the server and the …
N5 · Complete systemTurns the kilo you bought into real portions and Saturday's demand into a purchase list, using a yield factor …
Included in the always-growing library
Access to every published assistant, adapted to your AI and personalised with your restaurant's data, updates included.