I will forecast your sales and optimise stock levels

Running out of your best sellers while other products gather dust? Most shops reorder with "last month's sales plus a safety margin". I replace that with a weekly forecast per product and order quantities that balance missed sales against the cost of stock.

On a real UK online wholesaler data (748 products, six months not used for development), this may cut the cost of missed sales and stock by 21 % against the rule of thumb, while serving more customers, according to simulation. See the live demo: https://alb-i.github.io/demand-showcase/demo/

How it works:

1. I clean your sales export (returns, cancellations, one-off bulk orders).
2. I forecast every product's demand for the next weeks, with a realistic range.
3. I turn the range into an order list for your delivery times and margins, and simulate it on your own history against your current rule, so you see the difference in your currency before you change anything.

You get the results as CSV files, the charts, and a documented Jupyter notebook that reproduces them. No lock-in.