Walk the plot and describe what you find, the way you would say it to a client. Upload the footage. Every defect comes back as its own entry, worded from your own description, pinned to the second you said it, with the clearest frame from the video attached. You check it, correct anything you want, and send it.
Spoken on site, mid-walk. No form, no typing.
Transcribed word by word against the clock.
Its own entry: room, category, severity, timestamp.
On your letterhead, with the clip one tap away.
The only thing that changes about your survey is that you say it out loud instead of writing it down.
Walk the plot with your phone and describe what you find, the way you would say it to a client. No app to learn, no forms to fill on site, no kit beyond the phone in your pocket. Film a single room or a whole handover in one go, and upload as many clips as one property needs.
Processing runs on our side, not yours. Close the laptop, get in the van — the report is being built whether the tab is open or not.
Your narration is transcribed and timed word by word, so the system does not just know what you said — it knows exactly when you said it. That timing is what lets every defect point back at the right moment in the footage instead of a vague guess.
Each defect is pulled out and written up on its own: what is wrong, which room it is in, and how severe it is. Nothing is invented, because it is your description, structured. One sentence about two faults becomes two entries, and a passing remark that was never a defect does not become one.
For each snag, Smart Snag grabs a burst of frames around the moment you spoke about it, throws away anything blurred or caught mid-pan, groups the near-identical ones together, and keeps only frames that genuinely show the defect. If the footage never caught it cleanly, you get no still rather than a misleading one — a wrong photograph on a snag report is worse than none.
What lands in your client's inbox is your document, on your letterhead, with evidence behind every line of it.
Every snag arrives with its clip cued to the right second, its stills, your notes, the room, the category, the severity and a status you can move as work gets done. Filter by any of it, mark up an image to point at the exact crack, or add your own photos.
Produce a branded PDF for the client, or share a digital report they can open on a phone with the clips playing inline. Each snag can carry a link straight to the video of that defect, so the person who has to fix it can watch it happen rather than argue about it.
Transcription mishears things. A room gets guessed from context. This is the part most tools go quiet about, so here it is plainly: the machine does the typing, and you keep the judgement.
Anything the system inferred rather than heard is marked AI-estimated — pending inspector verification. It never reaches a client's report dressed up as something you confirmed.
Change the room, the category, the severity or the wording in place. Take a snag out of the report entirely if it was never a defect. Removed snags stay out of every export and every share link.
No report exports, sends or publishes on its own. The document that carries your name and your professional liability leaves only when you decide it is right.
The survey was never the slow bit. The slow bit was afterwards — typing every defect up from memory and scribbled notes, sorting photos into rooms, working out which picture went with which crack, formatting it into something a client would accept. That is the part Smart Snag does while you are driving home.
Film one room, upload it, and watch the report build itself.
No card and no install. Every snag is on screen before we ask for an email address.