Multi-Site Maintenance Request Portal

The problem this kills

You run facilities for twenty, fifty, two hundred locations. A leak at the Riverside store goes to your inbox. A broken freezer at the downtown branch goes to your cell. A flickering light gets reported in a group chat and forgotten. Every site has a different local vendor, every category has a different urgency, and you're the human switchboard trying to remember which plumber covers which city and whether you already dispatched someone for that exact thing yesterday.

Meanwhile nobody can answer the questions that actually matter: Which sites are blowing their SLA? Which vendor is slow? How much have we spent at the Eastgate location this quarter, and did anyone approve that $4,000 HVAC call? The requests live in email, text, and memory - so duplicates get dispatched twice, urgent things wait while trivial things jump the line, and there's no record when a regional manager asks why the Northpark store has had a broken door for three weeks.

This tool replaces the switchboard-in-your-head with a real portal: staff submit, the system routes to the right local vendor automatically, you approve the dispatch, and an SLA clock and per-site dashboard keep everyone honest.

What you'll build

A small internal web app, just for your organization, that:

• Lets site staff submit a maintenance request from a simple form - their site, the asset, a category, urgency, a photo - seeing only their own site's requests.
• Auto-routes each request to that site's assigned local vendor for that category, using your site/vendor/SLA list.
• Dedupes so the same problem at the same site (same site + asset + an already-open request) can't be filed or dispatched twice.
• Puts every request in front of the central coordinator, who reviews and approves dispatch - and gets a clear flag when the estimated spend is over your cap so big jobs get extra eyes.
• Starts an SLA clock the moment a request is dispatched, using that site's SLA target, and tracks status (submitted -> approved -> dispatched -> in progress -> completed).
• Shows a per-site SLA dashboard: open vs. closed, on-time vs. breached, average response time, spend by site and vendor.
• Exports dispatched work orders and the full request log as CSV, in the columns your existing systems expect.

What's inside the Implementation Plan

The plan is a single markdown file you paste into Claude Code (a free AI coding agent). It walks the agent through building the whole tool, step by step, each step ending with a ready-to-paste prompt.

The most important part: the plan opens by interviewing you about your business. Before it writes a single line, the agent asks how requests reach you today, how your sites and vendors are organized, the real names and codes for your sites and asset categories, your typical and peak request volumes, your per-site SLA terms, your spend cap, and your messiest edge cases (a site with no assigned vendor, an emergency after hours, a job that needs a quote first). It reads a short tailored spec back to you, you confirm it, and only then does it build - so you get a portal shaped to how your locations actually run, not a generic template you have to fight.

Inside you'll find:

• The discovery interview and how the agent turns your answers into the data model.
• The full build: database, login, the site-staff request form, the auto-routing engine, the coordinator's approval-and-dispatch queue, the SLA clock, the per-site dashboard, and the emails.
• The hard human dispatch gate and the over-cap spend flag.
• Verification steps so you can prove it works, and the CSV-export fallback so it's fully usable even before you connect it to your CMMS or finance system.

The governance it includes (this is the point)

This isn't a toy. The plan builds in the controls a multi-site facilities operation actually needs:

• Login so only your team and your site staff can see or touch anything.
• Row-level security so each site sees only its own requests - never another location's - while the central team sees everything in your organization.
• A complete audit trail - every submission, routing decision, approval, dispatch, status change, and reassignment is logged with who and when.
• A hard human-in-the-loop gate - the AI routes and drafts the work order, but a request becomes a dispatched work order only after the coordinator approves it; nothing is auto-dispatched.
• An over-cap spend flag so any job above your threshold is held for explicit approval before it goes out.
• Duplicate guards so the same open issue at the same site can't be dispatched twice.

Who it's for

Multi-site facilities managers, regional facilities coordinators, and retail or franchise operations teams who are juggling maintenance across many locations from their inbox - and who want a real, auditable portal with per-site SLAs without hiring a developer or buying a heavyweight CMMS. You don't need to write code. You need your site/vendor list, your categories, your staff-per-site sheet, and a weekend.

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