Real-Time On-Hand & Availability Dashboard

The problem this kills

Right now, when a customer asks "can you ship 40 of these today?", somebody guesses. They eyeball a stale spreadsheet, mentally subtract the orders they think are already promised, maybe ping the warehouse, and quote a number that's wrong often enough to cause oversells, scrambles, and apologies.

The real "available" number is never just on-hand. It's on-hand minus what's already committed to open orders, sometimes plus what's inbound and arriving soon. That math lives in someone's head or in a tangle of VLOOKUPs that nobody trusts. So sales under-promises and loses deals, or over-promises and creates a fire.

This kills the guessing. One screen, one trusted number, refreshed on a schedule, and signed off by the person who actually owns inventory accuracy.

What you'll build

A clean internal web app where your team logs in and sees, for every SKU at every location: on-hand, allocated/committed, in-transit, and a single available-to-promise (ATP) number computed the way your business defines it. Fast search and filters by SKU, location, and stock status. A clear "last refreshed" timestamp so nobody quotes from yesterday's data.

Crucially, a new data load doesn't go live automatically. It lands as a pending snapshot that your inventory control lead reviews, corrects if needed, and approves. Only then does it become the published number teams quote from. Everything is logged: who imported, who approved, what changed, and when.

What's inside the Implementation Plan

• It starts by interviewing you about your business. Before building anything, the plan has the AI agent ask you how you actually work - your SKU and location naming, where on-hand and allocations come from, how you define "available," whether inbound counts, your peak volumes, and your messy edge cases. It reflects a short tailored spec back to you and waits for your thumbs-up. You get a tool shaped to your warehouse, not a generic template.
• A step-by-step build you complete by pasting ready-made prompts into Claude Code.
• The exact data model for SKUs, locations, on-hand, allocations, in-transit, and approved snapshots.
• A configurable ATP formula you set with the agent during the interview.
• The import flow (CSV/Google Sheet today, an API later) with duplicate guards on SKU + location + snapshot.
• The approval gate, audit trail, and a clean CSV export in the columns your other systems expect.

The governance it includes (this is the point)

• Login so only your team can open the dashboard.
• Row-level security so each organization only ever sees its own data.
• A human approval gate: imports land as a pending snapshot; the inventory lead reviews and approves before the numbers go live and become quotable.
• A complete audit trail: who imported, who approved, what manual corrections were made, and when.
• Duplicate guards so the same SKU/location can't be double-counted within a snapshot and the same file can't be loaded twice.

Who it's for

Customer service reps who quote availability, salespeople who need a trustworthy number on a call, inventory control specialists who own accuracy, and ops managers who are tired of oversells. If you live in a spreadsheet today and wish it were a real, shared, trustworthy screen - this is for you. No coding background required.

You've got this. Paste the first prompt and let the agent interview you.