A shop calendar that knows what will run out this week

Description and use

Shop Stock Predictor uses TabPFN to forecast weekly sales and Gemma 4 to rewrite notes, helping a small shop owner determine which items will run out and how much to reorder.

Problem

The shop owner manually decides reorder quantities by walking shelves, often ordering too much or missing items that run out early.

Outcome

On three months of real bills, TabPFN achieved a per‑day error of 0.63 versus 0.81 for plain average and 0.80 for same‑weekday average, and a per‑week error of 3.27 versus 3.71 and 3.70.

Workflow

• Upload sales report
• Upload stock report
• TabPFN predicts daily sales for each item
• Plain arithmetic determines run‑out day and order quantity
• Gemma 4 rewrites each line of the note
• Generate order list for WhatsApp or print
• Print receipt via Bluetooth printer

Evidence

He gives it those two reports, and it says what will run out in the next 7 days and how much to order.

Source

On three months of real bills from the shop, over six hidden weeks, in pieces off per item: Plain average 3.71, Same‑weekday average 3.70, TabPFN 3.27.

Source

The app has an honesty test. It hides the last week of sales, predicts it, and shows how far off TabPFN was, next to two averages that need no AI.

Source