AI assistant · N3 Analysis with a verdict
Strategic Diagnosis Builder for Restaurants
It examines your restaurant area by area and hands you a diagnosis report: the findings table with evidence, severity and profit impact, the main bottleneck identified, and a four-week work prescription.
N3Analysis with a verdict
Crosses YOUR numbers and takes a stand, with a table, units and currency. Not options: a recommendation you can defend.
◆ This assistant 24 of the 164 published assistants are at this levelIs this for you?
What it does and what you get
- When the business is not producing the profit it should, the owner's instinct is to attack the loudest symptom: cut prices, buy more ads, replace the cook.
- This assistant does what a consultant does on a first visit: review the business areas in order, separate symptoms from causes, and tell you frankly which problem, solved first, unblocks the rest.
- You leave the session with a diagnostic report and a work prescription for the coming weeks.
What you end up with
A diagnostic report with a findings table by area (evidence, severity, profit effect), the identification of the main bottleneck, and a 4-week work prescription with actions and their rationale.
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: «We sell more than last year and profit will not move. Audit the whole business and tell me what is really failing.»
| Area | Finding | Evidence | Severity | Profit lost per month |
|---|---|---|---|---|
| Profitability | The 12 highest-volume items have no written portion weights | Weighed in service on 3 plates: the Bandeja de la casa runs 14% over its costing | High | 2,600,000 |
| Operations | Location 2 runs lunch staffing during low-occupancy bands | Sales by time band, last 8 weeks | High | 4,100,000 |
| Sales and channel | The delivery platform carries the dining-room price with the commission on top | Platform settlements, 8 weeks | High | 1,700,000 |
| Value proposition | The menu describes dishes in one line; "home cooking, served the way a restaurant serves" appears nowhere | Reading of the menu, the delivery channel and the Instagram profile | Medium | 1,900,000 |
| Team | 5 departures in 12 months across 24 people; replacements start with no written recipe | Payroll and onboarding log | Medium | 1,400,000 |
Total: 11,700,000 a month, close to 5.5 points of margin on sales of 214,000,000. The audit's four remaining findings together add up to less than the smallest of these five.
SUPUESTO: profit lost was calculated with the contribution margin implied by the declared prime cost (62%). If the real margin were 5 points lower, every figure in the last column drops proportionally and the order of the findings does not change, but the prescription goes from 4 weeks to 6.
The bottleneck is not the delivery platform. It is that no standard recipe and no per-plate costing exist.
The chain, link by link: with no written portion weights, food cost drifts on its own and nobody knows why → with no reliable cost per plate you cannot set a separate platform price or evaluate a new line → every incoming cook learns by ear, so waste climbs and the year's 5 departures multiply it → and the owner compensates for the profit that never shows up by selling more. That is why sales rise and profit sits still: the business is growing on top of a cost nobody controls.
What happens in 90 days if nobody touches it: with sales growing at the pace of the last 12 months and prime cost frozen at 62%, the goal of moving operating margin from 8% to 14% stays six full points away and the year closes around 8%. More sales, same profit, more exhaustion.
The uncomfortable part, said straight: the problem the owner mentions first —the platform commission— is third in money and first in noise. Fixing it before the portion weights leaves the commission neatly calculated on a cost that is still false.
Illustrative example generated with a fictional restaurant. Not a promise of results, not a client case.
This is 2 of 3 parts. Behind the wall is **the 4-week prescription**: one action per week with an owner and the figure that confirms it, starting at the bottleneck and not at the easy part, plus all nine findings on their severity-versus-effort matrix.
See the full example — free accountWorks with these AIs
How to use it
Which of your data it uses
This assistant works with 11 of the 23 fields in «My Restaurant», among them Restaurant name, Business type and Years in business. 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.
v1.3.02026-08-27It was called «Consultor Estrategico para Restaurantes». The name said who the AI was pretending to be, not what the owner takes away. Diego's instruction (2026-08-27): «what worries me most is that the client understands what they are for, because the names do not say by themselves what they do». The title now names the DELIVERABLE and the summary lists it. Not one comma of the assistant's body changes: what changes is what is read before opening it.v1.2.12026-08-26Declares what it does NOT do. `motor/alcance.py` had flagged it as high risk —its title names a role, or its body promises to act on a system— and with no `limites_es`/`limites_en` the public page was promising by omission.v1.2.02026-08-20Rooting. The sheet named the house in the header and then audited any restaurant at all. The four working sections were rewritten so the data decides: the interview asks which of their strong dayparts hurts, the sales mix is audited on their star dishes and each channel on its own, the prescription never asks for more than their team can execute, a finding that would fit the place next door is thrown out, and the next step starts with whatever moves their goal most. Nine fields at work, seven inside the TASK. Text was swapped, not added: the body went DOWN from 800 to 790 words and the duel holds at 100.0.v1.1.02026-08-13Rewritten under the margin mandate: 6 numbered audit steps, a priced finding with evidence per area, 5 named output blocks with their columns, 8 business data points, consulting-room warnings and figures with units. Duel 51.6 -> 100.0.v1.0.02026-08-09Version inicial de la biblioteca viva.Same method, another variant
Inside the app this is a single control: you press it and the assistant reframes itself, including the variants that don't appear here because they don't have their own page yet.
More assistants in this category
You choose what you put on the table —a business decision, a growth opportunity or a new business line— and ou…
N3 · Analysis with a verdictTakes everything on your plate, weighs it by its effect on profit, and returns three defensible priorities, a …
N4 · Executable planDesigns your restaurant's complete strategy on a single coherent page: who it exists for, how it intends to wi…
Included in the always-growing library
Access to every published assistant, adapted to your AI and personalised with your restaurant's data, updates included.
