Capture any phone
Point, shoot, send — from the phone that's already in the crew's hand. No app to install, no forms to fill, nothing new to learn.
Photo-to-record, in one snap
Your AIcan see.
Snap a photo — a job site, a menu, a receipt, a storefront — and SIGNL knows what it's looking at and files it where it belongs. Your phone camera becomes an intake device.
Live on real job sites, not a mockup
✓ Any phone, no app to install
SIGNL VISION
CAMERA
READING
reads Job-site · framing 0.98
filed to Project ▸ Framing · Day 3
on the record — findable, timestamped FILED
A photo is the shortest path from the field to the record — and it's the one input everyone already knows how to send.
What the camera does now
Snap it, and it does the filing for you.
Classify it knows what it sees
SIGNL reads the photo and names what it is — a job-site shot, a menu, a receipt, a storefront — the way a person would at a glance, not a blank upload.
File where it belongs
Because it knows what it's looking at, it routes the photo to the right place on its own — the project, the menu, the record — instead of a camera roll no one revisits.
On the record findable
Every photo lands timestamped and searchable, tied to the job it belongs to. The camera roll becomes a record you can actually pull from later.
How it works
Snap it. It reads it. It's filed.
01 Snap
You take the photo.
Out in the field, at the counter, on the drive back — one photo from any phone. That's the whole ask on your side.
02 It reads
Vision sees what it is.
SIGNL looks at the image and classifies it on sight — what kind of photo, and which job or place it belongs to — no tagging by hand.
03 It's filed
It lands where it belongs.
The photo is routed and filed automatically, on the record and ready to find — before you've put the phone back in your pocket.
and the next photo files itself the same way — the record keeps itself current
Receipts, not hype
Already reading real photos.
A construction crew LIVE
22 job-site photos
Captured, classified, and filed to the right project. The crew's camera roll turned into the project's photo record — sorted on the way in, without anyone tagging a single shot.
A restaurant LIVE
Menu read on sight
Photos of the menu get classified the moment they land. A picture from the kitchen's phone is recognized for what it is — no re-typing the menu to get it into the system.
Real, live tenants on the same rails you'd run on. Names withheld on purpose — your photos are your business, not our marketing. We'll show you a working one on a call.
What's coming — and what isn't here yet
We'll tell you what it can't do yet.
Read the scope, draft the estimate.
Today Vision tells you what a photo is and files it. Reading a job-site photo to work out the scope and a first-draft estimate is where we're headed next — it isn't built, and we won't pretend it is.
Your own eyes, on your own hardware.
Right now Vision runs on a frontier vision model. Running it on your own machine is the plan — but our in-house model reads at about 87.5% on a real-world set, under the accuracy bar we set before we'd trust it with your business. When it clears the bar, it flips — not before.
Live on-site cameras.
Today Vision reads a photo you send. Extending it to live on-site cameras — Vision watching the floor, the counter, the line as it happens, not just reading a snapshot — is where it's headed. It isn't live yet, and we won't pretend it is.
Straight about today: capture, classify, and file are live and running on real photos. The three above are where we're going — listed here so the line between shipped and coming stays honest.
Point a camera at it
See it read a photo.
Open the live portal and watch a photo get read and filed, or book a call and we'll stand yours up on the rails that are already running.
portal.signloffice.com
signl@ricreateagency.com