AI for retail, built for hardware and building-supply stores.
We encode the rules your team carries in their heads — pricing exceptions, count tolerances, tag-labor thresholds — inside the systems you already run.
Official services partner of the platforms defining AI
NVIDIA
Anthropic
Margin leaks one price change at a time.
From owners, ops managers, and pricing leads inside independent hardware co-ops and regional chains — all running Epicor Eagle-class POS/ERP as the system of record.
The weekly Excel everyone actually reads
Built by hand and emailed to the org — the thing everyone actually reads: 5,700 price changes across 19 stores, no clear view of margin gained versus lost.
Prices set somewhere else
The co-op sets half the prices — no database against the big boxes, just the wholesaler’s ~1.15x baseline — and one contractor can be a third of a store’s business.
Markdowns that swallow promos
Clearance markdowns overwrite advertised promos under one ERP markdown flag, and underwater SKUs — replacement cost below weighted-average cost — stay invisible without four cost columns side by side.
Counts nobody trusts
Count error rates run 30–40% with variances hand-keyed from paper into the PIP file, and inventory accuracy sits near 75% against a 95% best-in-class bar.
Tags faster than hands
Changes stop at zero on hand, but tags still queue up; special orders never picked up become hundreds of thousands of dollars of product no one will buy.
The workflows worth encoding first.
Each card is a workflow or capability built on the same pattern: your pricing exceptions, count tolerances, and tag-labor thresholds — encoded inside the systems you already run.
The pricing summary goes out every week, built by hand in Excel and emailed to the org. It is the thing everyone actually reads.
Led by Peter Enestrom.
Asked by owners, ops managers, and pricing leads.
The straight answers, before you book anything.
It reads the exports your POS and ERP already produce — cost changes, price files, count sheets, receiving logs — and applies your rules to every line: pricing exceptions, count tolerances, tag-labor thresholds, receiving cutoffs. What the rules resolve happens on its own; what they cannot lands in a review queue for a person. Nothing on the sales floor changes.
Yes. The co-op file is one input, not the whole pricing picture. Contractor accounts, clearance timing, tag labor, count tolerances, and receiving discipline are store-side rules nobody wrote down — and that is where the margin leaks. The system still flags what co-op cost changes do to your margin, even when you cannot override the retail.
No. The rule layer runs on exports from the Epicor Eagle-class system you already run and pushes back into it. It is not a POS replacement and not a rip-and-replace — your team keeps its screens, and the exceptions come to them.
The platform assumes your data is clean and your rules are generic. Yours aren’t. Optimization suites price for the enterprise and assume an enterprise team to run them; POS and ERP vendors sell modules, not your exceptions. We encode your exceptions inside the systems you already run.
Two weeks inside one store’s pricing and count flow, fixed fee. We map the rules, quantify the leak in your own numbers, and hand you a ranked opportunity map — the map is yours whether or not you continue. From there the build encodes the rules, and operations become exception-only: people see only what the rules cannot resolve, and the weekly Excel report writes itself.
Start with one workflow.
Tell us where your team loses hours. We will come back with a straight answer on whether AI can help, what it would take, and what it would pay.