Practical guide
AI fluency for supply chain analysts: check stock timing
Trace a fictional item balance across demand, confirmed receipts and delayed arrivals without treating total supply as timely supply.
Supply chain analysts can use AI to organize a time-phased balance while checking when stock becomes usable. This fictional planning exercise separates opening stock, demand and confirmed receipts. It focuses on a timing shortfall, not supplier scoring, a real purchasing recommendation or an optimized inventory policy.
Fix the item, location and time boundary
A stock balance is meaningful only when its quantities refer to the same thing.
The exercise uses one fictional item at one location. Opening usable stock on Monday morning is twelve units. Monday demand is eight, Tuesday demand is seven and a confirmed receipt of ten units arrives Wednesday morning. Ignore other locations and substitutions unless the case explicitly introduces them.
Ask AI to build a day-by-day table, naming when receipts occur relative to demand. Do not combine stock held elsewhere with locally usable stock. Keep quantities in individual units; a supplier carton is not interchangeable unless the number of units per carton is supplied and checked.
Calculate the balance in time order
An eventual surplus does not remove an earlier shortage.
After Monday demand, four units remain. Tuesday demand exceeds that balance by three units. Wednesday’s ten-unit receipt can cover those three units of unmet demand, leaving seven units if there is no additional demand. State whether the table represents unmet demand as a negative balance or as a separate backlog.
Do not describe the opening twelve plus the receipt of ten as enough for all fifteen units on time. That total ignores the Wednesday arrival. Ask AI to explain the earliest shortfall in a sentence and check that its narrative agrees with the table rather than only the final total.
Keep receipt certainty separate from the arithmetic
A balanced spreadsheet cannot turn an unconfirmed arrival into usable stock.
Change the receipt from confirmed Wednesday to an unconfirmed Wednesday estimate. Keep the numerical scenario but label the assumption. The supported action is to request arrival confirmation and examine the consequence of delay, not to promise fulfillment based on a convenient date.
For a sensitivity case, move the receipt to Thursday. The shortfall persists for another day; the eventual quantity is unchanged. Do not invent a probability for the delay. State which operational decision depends on the arrival date and who can verify it in the fictional packet.
Compare responses without making unauthorized commitments
A recovery option needs its own evidence and decision owner.
The exercise offers three possible responses: transfer stock from another location, agree a later fulfillment date or investigate an earlier receipt. None is available merely because AI lists it. Ask what quantity, timing and authority would need confirmation for each option before relying on it.
If a transfer is proposed, check that its transit time addresses Tuesday’s shortfall rather than arriving after the same receipt. If a later date is proposed, distinguish a suggestion from an accepted customer arrangement. The task is to prepare a reviewable choice, not execute purchasing or delivery actions.
Deliver the time-phased exception note
The handoff should make the earliest problem and its unresolved dependency easy to find.
Include the item and location, opening stock, daily demand, receipt timing, earliest shortfall and alternative-date scenario. Show the calculations so another reader can reproduce them. Keep confirmed facts, planning assumptions and possible responses in different fields instead of mixing them into a single forecast sentence.
Introduce a carton conversion of five units per carton and ask the worker to translate a two-carton receipt without changing the ten-unit result. Review timing and unit discipline within the case. This differs from an order exception queue, whose main challenge is reconciling individual records and ownership.
Sources and scope
NIST: AI risk management. Skills England: workplace AI foundations.
These references provide background, not validation or endorsement of this exercise. The case details, calculations and suggested review questions are original instructional material. Use them to discuss observable work, not to infer customer outcomes, professional credentials or performance in every setting. Before adapting the exercise, confirm the relevant facts, approved tools, data permissions and decision owners. If you change the case, revisit the expected answers and checks as well. These examples describe practice tasks, not a promise that a particular product includes the fictional features.
Sources: [1] [2]
Sources
- 1.AI RMF Core · NIST
- 2.AI foundation skills for work benchmark · Skills England