AI inspections for Lumiform

Overview

A completed Lumiform form is a well-built record of the questions you asked and the evidence you asked for; the photos inside it hold a good deal more than that. Lumiform gives a form a Photo response type, photo comments, form attachments, negative-response marking, auto-triggered actions and a report generated the moment the form is completed, with the images in it. RapidEye reads those same images against the last known state of the same area at that site, and returns one finding per problem, whether or not a question in the form covered it.

How it works with Lumiform

Lumiform's shape is a form built once and run many times, so the useful unit here is not a single submission but a series of them: the same site, the same template, week after week. That is what makes a photo archive worth reading.

• Opening and closing checks. The line, the prep bench, the walk-in, the restrooms, photographed at the same two points of every day. A finding comes back when something is different from last time and nobody said so: a cracked shelf liner, a wall tile gone, a bin lid that no longer closes, a light out over the pass.
• Photo answers on the questions you already ask. The photograph proving the sanitiser bucket was changed also shows the floor, the skirting, the cable run and the corner of the counter. One frame, one question answered, and a good deal more information sitting in it unread.
• Negative responses and the actions they trigger. Lumiform lets you mark response options as negative and fire an action automatically when one is chosen. That handles the failures your template predicted. RapidEye covers the other case: nothing was marked negative, every answer was fine, and the photograph still shows a problem.
• Action closure proof. An action in Lumiform carries a title, priority, timeline, assignee and visual proof. RapidEye reads the closure photo against the photo the action was raised from, so a repair is compared rather than asserted, and one that did not hold shows up on the next visit.
• Entities as the unit of comparison. Lumiform's entities cover locations, vehicles and machinery, and forms and actions get assigned against them. RapidEye keeps its baseline the same way, per site and per area, so findings arrive with the location already attached.
• Multi-site comparison. When forty sites run the same template, the interesting question is which sites are drifting and which are holding. Reading the photographs gives you that in terms of physical condition rather than completion rate.

A finding names the site and the area, where in the frame the problem is, what kind it is (damage, cleaning, maintenance, or something missing), how bad it looks, the two photographs that show it, and a suggested next step. Each one is filed in the RapidEye dashboard, mirrored to Slack or email when routed. What happens next in Lumiform is yours to decide: an action with an owner, a note to the franchisee, a line in the next audit, or nothing. Photos are the usual input rather than the only one, so teams that film a walkthrough of a site, that hold 3D scans of a fit-out, or that want findings scored against a standard of their own, get that set up with us on the first call.

How to connect Lumiform to RapidEye

1. The sites you already inspect. Your locations are set up as entities in Lumiform with the forms you already run scheduled against them. Whatever is assigned today is what RapidEye reads, and no template gets rebuilt.
2. Photos on the answers. Use the Photo response type, or a photo comment, on the questions where the picture is the point, and mark the field required where it matters. If a form carries no photos today, a RapidEye capture link goes out instead, opening on any phone with nothing to install and asking for a fixed set of angles per area.
3. Add RapidEye as a user. In Lumiform Desktop, go to Administration, then Users, then Add User, and add Rapid Eye with a role that can view all forms. The modal has the exact fields.

That is the whole setup. Frontline teams open the same app and fill the same forms, and the first findings follow the first submissions.

How a Lumiform form becomes a record

Checked against Lumiform's help centre and product pages, September 2026.

Lumiform, made by Lumiform GmbH, describes itself as a frontline operations platform for creating, managing and documenting checklist, audit and inspection processes. According to Lumiform's help centre documentation (lumiformapp.com), a form is started from a template in either the web app or the mobile app, worked through question by question with required items marked by an asterisk, and completed with a Complete form step that then offers to assign all the actions raised along the way before the report is saved or exported. The form builder behind it offers Multiple Choice, Text, Number, Signature, Photo and QR Code response types, plus if-then logic, scored answers, repeatable sections, response sets, negative-response marking and categories.

The four places a photo can live in a Lumiform form

This matters more than it sounds, because it decides what a photograph is attached to, and therefore what it can later be compared against.

Photo locations as documented by Lumiform. The bottom line of each tile is what RapidEye does with that photo.

What a submitted form leaves behind

Lumiform generates the report automatically on completion, which means the archive builds itself whether or not anyone goes looking. The pieces of it, and what each one settles:

Does Lumiform already check the photos my team takes?

Yes, at the moment of capture. Lumiform documents an AI photo validation feature: the team member takes a photo, AI checks whether it meets your standards, and they get compliance feedback on the spot so the issue can be corrected while they are still standing there. That is genuinely the right place for that check, and it is the sort of thing that stops a bad answer from being submitted at all. Alongside it Lumiform documents AI that counts items in an image, blurs faces and private information for safe reporting, scans documents, captures text with OCR and generates descriptions for uploaded images.

RapidEye is a different pass over the same pictures, and the difference is time. A capture-moment check evaluates one photograph against the standard for the question being answered. RapidEye holds a baseline for each area at each site and compares across visits, so what it reports is change: the thing that was not there in March and is there in September, the item that used to be in the corner and is not now, the wear that nobody would flag on any single day but that is obvious across six months of the same shot. Neither reading replaces the other, and the second one only becomes possible because Lumiform has already done the hard part, which is getting a frontline team to photograph the same places, in the same order, at every site, offline or not, week after week. Everything RapidEye does here is downstream of that archive existing.

Lumiform integration FAQ