# Future-State Analysis

## Basis and Desired Outcome

The prototype models a future commercial-kitchen operating approach that addresses gaps in the current recipe collection and menu-planning environment.

User-provided future-state examples:

- Proper mass-to-volume conversion.
- Unconventional commercial-kitchen volume units and conversions, including hotel pans.
- Scaling from sub-recipe ingredients through recipe rollups.
- Forecasting tied to production records.
- Automated ordering forecasts.
- Analysis of forecast-to-production accuracy.

## Target Operating Model

| Future-State Area | Target Capability | Prototype Evidence | Evidence Quality |
| --- | --- | --- | --- |
| Recipe measurement authority | Recipes and base foods carry enough mass/volume authority data to support cross-family scaling. | `src/services/unit_conversion_service.py`; `src/services/recipe_service.py`; recipe scaling tests | Observed/User-provided |
| Hotel-pan production units | Commercial-kitchen units such as hotel pans are supported in scaling and Forecasting where operators need them. | `src/config/hotel_pan_units.py`; `src/config/units.py`; forecast/advanced scaling tests | Observed/User-provided |
| Sub-recipe rollups | Nested recipes can be viewed hierarchically or flattened so ingredient demand can roll up accurately. | `src/services/recipe_flattening_service.py`; `src/services/recipe_scaling_service.py` | Observed/User-provided |
| Menu-to-forecast workflow | Menu assignments become forecastable production plans by service date, meal period, and concept. | `src/services/menu_service.py`; `src/services/menu_forecast_service.py`; menu docs/tests | Observed |
| Forecast-to-production tie-in | Forecast summaries become production record snapshots with actuals, variance, implied demand, and posted facts. | `src/services/production_record_service.py`; `docs/agent_context/menu_builder.md` | Observed/User-provided |
| Ordering forecast | Inventory planning converts menu/recipe demand and current on-hand data into shortage/reorder signals. | `src/services/inventory_usage_service.py`; `src/services/inventory_ordering_service.py`; inventory roadmap | Observed/User-provided |
| Forecast accuracy analysis | Production history and item trends analyze actual production outcomes against forecast expectations. | `src/services/production_record_service.py`; recent Git history; README roadmap | Observed/User-provided |
| Vendor and invoice bridge | Vendor catalog/invoice data supports future ordering, costing, substitutions, and item-specific sourcing. | `database/schema.sql`; `src/services/inventory_bridge_service.py`; `src/services/inventory_invoice_service.py` | Observed |
| Analytics readiness | Analytics consumes stable menu, usage, inventory, purchasing, cost, and production-variance contracts after operational workflows settle. | `README.md`; `AGENT.md` | Observed |

## Future-State Process

```mermaid
flowchart LR
    A[Recipe Truth with Measurement Authority] --> B[Sub-Recipe Rollups]
    B --> C[Menu Builder]
    C --> D[Forecasting with Kitchen Units]
    D --> E[Production Record Actuals]
    E --> F[Forecast Accuracy Analysis]
    D --> G[Ingredient Demand]
    G --> H[Inventory On Hand]
    H --> I[Ordering Forecast]
    I --> J[Vendor / Invoice / Catalog Bridge]
    F --> K[Future Analytics]
    J --> K
```

## Success Measures

- User-provided/Inferred: Operators can scale recipes across mass, volume, each, and hotel-pan contexts without manual spreadsheet work.
- User-provided/Inferred: Nested recipes roll up to ingredient needs so production and inventory demand are not understated.
- User-provided/Inferred: Forecasts become production records without duplicate entry, and production records produce actual-vs-forecast learning.
- User-provided/Inferred: Ordering forecasts are generated from menu/recipe demand, current on hand, vendor preferences, and catalog/invoice data.
- Inferred: Analytics begins only after the operational facts are stable enough to trust.

## Future-State Guardrails

- Observed: Hotel-pan and other advanced units should be used in scaling/forecasting contexts, not blindly accepted in base recipe authoring fields.
- Observed: Temporary each conversion assumptions should not overwrite Recipe Collection item truth.
- Observed: Inventory planning should keep calculated need separate from rounded purchase suggestions until purchase rules are validated.
- Inferred: Automated ordering should remain explainable, with visible demand source, on-hand basis, vendor source, delivery timing, and conversion assumptions.
- Inferred: Forecast accuracy should distinguish draft operational memory from posted production facts.

## BA Alignment

- Activity areas: Analyze strategy; evaluate solution; analyze requirements and define design options.
- Techniques used: Future-state modeling, process modeling, document analysis, data modeling, acceptance criteria.