DECCANPRO — RESTAURANT INGREDIENT RECONCILIATION EXERCISE
https://deccanpro.com/guides/restaurant-inventory-vs-sales/

Fictional training quantities only. Not an import template, food recipe,
purchasing recommendation or proof of supported software features.
Use outside live service. Do not change production records for this exercise.

BASELINE — all milk quantities in ml
Opening: 2000
Received in the same period: 1000
Regular coffee: 6 drinks x 100 ml = 600 ml theoretical usage
Large coffee: 2 drinks x 150 ml = 300 ml theoretical usage
Recorded discard: 100
Expected closing = 2000 + 1000 - 600 - 300 - 100 = 2000 ml
Physical count: 1900 ml
Difference (counted minus expected) = 1900 - 2000 = -100 ml
Eight drinks is a dish count, not an ingredient quantity.
The -100 ml difference is unresolved; it does not identify a cause or person.

DELIBERATE CONFIGURATION ERROR — separate rehearsal
Map the large drink to 100 ml instead of 150 ml.
Drink count remains 8; theoretical usage becomes 800 ml.
Expected closing becomes 2100 ml with the same other movements.
Compare individual portion links, correct the test mapping and repeat.
Do not alter live recipes or use this deliberate error during service.

YOUR AGREED TEST
Date and exact cutoff:
Ingredient / base unit:
Pack size and conversion:
Regular portion / quantity:
Large portion / quantity:
Opening quantity:
Receipts:
Theoretical usage (show calculation):
Separate discards / other movements (state sign and reason):
Expected closing:
Count time and measured closing:
Counted minus expected:

OBSERVATIONS
Menu-to-recipe link demonstrated by / staff role:
Event that creates ingredient deduction:
Cancellation BEFORE preparation: observed bill and stock result:
Cancellation AFTER preparation: observed bill and stock result:
How already-used ingredients are represented:
Any transfer between stock locations:
Timing, units or missing movements to investigate:
Does any movement appear twice? Evidence:
Unsupported workflow / separate module or system needed:
Responsible person / next action / retest result:

Keep sample order and movement references, with private details removed.
Do not force a zero difference with an unexplained adjustment.