WhereHowSpace in ChatGPT — recorded demonstration
Captured November 3, 2026. Duration: 3 minutes 44 seconds.
Silent step-through screen capture of the real ChatGPT workflow; pauses shortened.
This is a synthetic planning example, not a customer facility or an engineering approval.

00:00 — Define the scenario
The user asks WhereHowSpace to plan an 80 × 100-foot rectangular warehouse with 30 feet of clear height, selective pallet racking, 144-inch aisles, 60-inch loaded pallets, no columns, a 60-foot maximum aisle run and no aisle tunnels. The user explicitly requests calculation and the layout.

00:05 — Calculate and view the result
ChatGPT invokes WhereHowSpace. The completed result shows 66 rack bays and 660 pallet positions. The layout is conceptual; building dimensions and rack requirements must be verified.

00:46 — Inspect the layout
The actual result is shown in 3D, then the same completed result is switched to a 2D floorplan at about 00:52.

00:59 — Edit without running
The user requests 156-inch aisles and asks to keep all other settings unchanged, without calculating yet. The settings are saved. ChatGPT explains that the displayed 144-inch-aisle result is now out of date and needs a new run.

01:48 — Explicitly run the revision
The user asks to run the updated scenario. The new visible result is 61 rack bays and 610 pallet positions.

02:22 — Repeat the unchanged request
The user asks to run the same unchanged scenario and check its status. ChatGPT confirms completion and reuse of the existing calculation, with 61 rack bays, 610 pallet positions and 156-inch aisles.

02:51 — Explain the free plugin's scope
The user asks whether the free plugin can run a Data Driven scenario from SKU and order data. ChatGPT says that the connected plugin does not support dataset import or Data Driven calculation. It distinguishes capacity estimates based on building and rack assumptions from the separate main web application's operational-data workflow.

Recording notes
The video retains the chronology of 620 genuine captured frames, with 21 pauses longer than two seconds shortened to two seconds. No synthetic interface, cursor, responses, narration or audio was added. The ChatGPT Activity popup is visible in parts of the recording; the final scope response is shown with it dismissed. This video does not demonstrate deletion or prove behavior during uncaptured gaps. Product results depend on the supplied assumptions.