Self-Serve Order Status Lookup
A customer-facing page where someone enters their order number and email and instantly sees their order's stage, ETA, and tracking link - read-only, fed by an ops-approved daily refresh - so your CSRs stop fielding "where's my order?" calls all day.
A customer enters their order number plus the email on the order and sees the order stage, an ETA, and a tracking link - safe fields only - and your status-call volume drops.
Before you start
- A list of your open orders with their current stage (a CSV or Google Sheet export is fine)
- Tracking numbers (or tracking URLs) for the orders that have shipped
- Free accounts: Vercel, Supabase, and Resend
- No coding experience required - you'll paste prompts into an AI agent
The problem this kills
Your customer service team spends half its day answering the same question: "Where's my order?" Every call and email is a person pulling up the same screen, reading off the same stage, and copy-pasting the same tracking link. It doesn't need a human - but today it gets one, every single time.
Meanwhile your customers hate it too. They don't want to call. They want to type their order number, hit a button, and see where their stuff is - the way every big retailer lets them.
This plan gives you that page. It's read-only and safe: it shows only the fields you approve, it never touches your real order system, and the data behind it refreshes on a schedule that an ops person signs off on before anything goes live to the public.
What you'll build
A clean, public lookup page:
- A customer types their order number and the email on the order, and clicks Look Up.
- They see the current stage (e.g., Received -> Picking -> Packed -> Shipped -> Delivered), an ETA, and a tracking link when one exists.
- They see only the fields you chose to expose. Internal notes, costs, margins, and supplier details stay hidden.
- Behind the scenes, your team uploads a fresh order/status export (CSV or Google Sheet) on a schedule. An ops person reviews and approves the refresh - and the public field mapping - before it replaces the live data.
- The page never writes to your system of record. It's a window, not a door.
What's inside the Implementation Plan
- It interviews you first. Before writing a single line, the plan has the AI agent ask you about your real order process, the exact column names in your export, your stage names, how you compute an ETA, and your messy edge cases - then it tailors the data model and the page to your business instead of dropping a generic template on you.
- A step-by-step build, where every step ends with a ready-to-paste prompt for your AI agent.
- The data model, the import/refresh flow, the public lookup page, and the approval gate - all spelled out.
- A "No API yet?" fallback baked in: pull the data straight from a Google Sheet or CSV on a schedule and you're fully live today, with no integration to your order system required.
- A verification checklist so you know it actually works before you share the link.
The governance it includes (this is the point)
- Login for your team. Only your staff can run refreshes and approve mappings; the public only ever sees the lookup page.
- Row-level security so each organization's data is walled off from every other - people only see their own.
- A full audit trail: who uploaded which refresh, who approved it, what field mapping was live, and when.
- A human-in-the-loop approval gate: the AI prepares the refresh and the field mapping, a person reviews exactly what customers will see, and only an explicit approval makes it live.
- Duplicate guards keyed on order ID, so the same order can't be double-loaded or shown twice.
- Safe-by-default fields: internal notes, costs, and margins are never exposed - the page shows only the columns you whitelisted.
Who it's for
CSRs and customer service managers drowning in status calls - and the customers who would much rather check for themselves. If you can fill in a spreadsheet, you can build this.
You've got this - paste the first prompt and let the agent interview you.