AI assistant · N5 Complete system
Interactive KPI Dashboard Builder for Restaurants
Pick the area above and whether you watch one location or the network, and the assistant builds THAT board: the full KPI sheet as an interactive file, every KPI with its formula, source, frequency and owner. It is the library's single dashboard builder: kitchen, service, purchasing, marketing, finance or management all come out of here with the selector.
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
- To stop managing with scattered numbers and hold ALL your restaurant's indicators on one interactive screen you filter by area, location and period.
- It differs from the Executive Dashboard in scope: that one is the owner's Monday view with 5 metrics; this is the full board — sales, costs, service, team — where every number has an owner and a definition, so no KPI is a number with nobody behind it.
What you end up with
A working, self-contained indicadores.html file: the full KPI board with filters by area, location and period, a traffic light per indicator against its target, and each KPI's card (formula, data source, frequency and owner by role); plus the KPI dictionary and the update routine.
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: «I want the full board: sales, costs, service and team for my two locations, with who answers for each number.»
| KPI | Formula | Source | Frequency | Owner (role) | Target | Traffic light |
|---|---|---|---|---|---|---|
| Food cost % | Consumption ÷ Food sales | Inventory + POS | Weekly | Head chef | The ceiling the house declares | ⬇️ lower is better |
| Prime cost % | (Consumption + Payroll) ÷ Sales | Inventory + payroll | Weekly | Manager | The ceiling the house declares | ⬇️ lower is better |
| Average ticket per channel | Sales ÷ Checks, dining room and delivery kept apart | POS | Daily | Manager | Quarter's target | ⬆️ higher is better |
| Table turnover | Services ÷ Available tables | POS + floor plan | Weekly | Floor manager | House range | 🎯 ideal range |
| Complaints per 100 covers | Complaints ÷ covers × 100 | Service log | Weekly | Floor manager | Quarter's target | ⬇️ lower is better |
SUPUESTO: no industry food cost or prime cost ceiling is hard-coded here: they shift by country, by format and by year. Until the house declares its own, the traffic light compares against the median of the sheet's own last 8 weeks, and says so on the KPI card. If the real ceiling were 2 points lower, the Centro location's food cost would go from green to amber.
Filters on top: `[All areas ▾] [Location: Centro ▾] [Channel: all ▾] [Period: Week 32 ▾]`.
| Area | KPI | Wk 32 | Target | Traffic light |
|---|---|---|---|---|
| Sales | Weekly sales | 21.4 M COP | 22.0 M | 🟡 |
| Sales | Average ticket (dining room) | $46,200 | $45,000 | 🟢 |
| Sales | Average ticket (delivery) | $38,900 | $42,000 | 🔴 |
| Costs | Food cost | 31.2% | declared ceiling | 🟢 |
| Costs | Prime cost | 56.8% | declared ceiling | 🔴 |
| Service | Complaints ×100 covers | 1.8 | 1.0 | 🟡 |
Every row opens on a click and shows the KPI card: formula, source, frequency, owner and its 8-week trend. The red prime cost explains itself on its card: August's extra staffing weighed on labor cost, and the note is dated.
The two tickets are kept apart on purpose. Averaging dining room and delivery into one row would have produced a «healthy» $43,400 ticket and hidden the channel that is losing. Same with the two locations: they are never averaged.
Bottom right, the seal: `Backup exported 25/08 · Safe merge active` — and in cells with no real figure, the grey `DEMO` label waiting for the data.
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 governance written inside the file** — who pastes which figure, when and where, the 15 min weekly reading with its order, the three-reds rule with its two-path decision, and the backup proof with the merge that never overwrites the newest data.
See the full example — free accountWorks with these AIs
How to use it
Which of your data it uses
This assistant works with 10 of the 23 fields in «My Restaurant», among them Restaurant name, Business type and Number of locations. 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.
v2.2.02026-08-27Becomes the HEAD of the dashboard family. Diego's brief: «take all this out, the c19 dashboard is going to be the integral one». The head sat on the Dining Reasons Dashboard, which is c05 and is just one more board; the area and location axes now live here, on the full KPI sheet. The card says so in its summary, because a control the card never mentions is a control nobody presses.v2.1.02026-08-25Body rewritten to pay down duelo debt: it scored 70.8 out of 90.0. The `de_a_una` signal read ZERO because of a design decision that overshot: the card orders questions «in a single block», which is right for the opening but left the rest of the work without a rule — the single block is kept for the opening, and it now adds that data missing mid-work is asked for one item at a time. `verificacion` read 2 of 5 because BEFORE YOU DELIVER was a row of questions with no checking verb, and `metodo` 4 of 6 because the four PHASES lived inside a paragraph: they are now six numbered steps at the start of the line. `arraigo` measured 0 fields in the work sections with ten «fields» declared; the TASK now ties cost targets to «food_cost» and «costo_laboral» and splits the ticket by «canales_venta». And the 32% food cost ceiling hard-coded as a house rule in the METHOD and repeated in the acceptance list is removed: the ceiling is declared by the restaurant, and where there is none, the AI writes its SUPUESTO: line and states what it is comparing against. Rises to 94.2.v2.0.02026-08-19The engineering standard from the master prompts (doc 21): non-negotiable role rules (zero invented data, garbage tolerance, frozen phases), a timestamped data contract, a three-type configurable traffic light with target-derived thresholds, a mandatory DEMO label, dated backups with a merge that never overwrites newer data, and the expanded acceptance checklist. Born from Diego's exercise: asking the built tool for the prompt that would build it.v1.0.02026-08-14Initial version. Born in the 🛠 tools line (artifact-titles batch 1) with the full lead-magnet kit and a sample from the fictional restaurant Sazón de Origen.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
The «sábana» —Latin American kitchen slang for the wide sheet where everything fits at once— is the master mea…
N5 · Complete systemThe PASS on one board: ticket times per dish against their standard, kitchen returns with fixed reasons, 86'd …
N5 · Complete systemPurchasing on one board: key-item prices against their history, supplier compliance measured at the scale, pur…
Included in the always-growing library
Access to every published assistant, adapted to your AI and personalised with your restaurant's data, updates included.
