← Back to Willifit.ai
AI-verified

How AI verification works

Every time you see an AI-verified badge on a parking garage, tunnel, or bridge, it means Claude Vision read the posted clearance sign from Google Street View imagery. Here's exactly what that pipeline does — and where we stop it from pretending to know things it doesn't.

The short version

1Street View fetch
2Claude Vision reads signs
3Confidence check
4Database update

For every parking garage in our database without a posted clearance, our batch pipeline pulls four Google Street View images at the entrance, asks Claude Vision what clearance sign is posted, and records the reading only if the model's self-reported confidence is high.

Step by step

1

Find the entrance

We start with the latitude/longitude already in our database (from OpenStreetMap, the FHWA National Bridge Inventory, or hand-curation). Before spending a cent, we ping Google's free Street View metadata endpoint — /streetview/metadata — to check whether imagery even exists at that point. If not, we skip and mark it "no imagery."

2

Pull four images

We fetch four Google Street View Static images at headings 0° / 90° / 180° / 270° (N/E/S/W) with a slight upward pitch to catch ceiling-mounted signage. Four angles, not one, because the clearance sign is usually visible from at most one or two directions — we want to give the model the best chance.

3

Ask Claude Vision, strictly

All four images go to Claude Vision in a single API call with a tightly-scoped prompt. The model returns JSON with:

  • found_sign — whether any clearance sign is visible
  • height_in — the posted height, in inches (null if not readable)
  • confidence — low, medium, or high
  • raw_text — exactly what the sign said
  • notes — occlusion, glare, or ambiguity flags

Key part of the prompt:

Things that COUNT as a clearance sign: * A sign showing a height in feet-inches ("11'6"", "13 FT 6 IN") * A ceiling-mounted clearance bar with a height number * A LOW CLEARANCE warning with a posted height Things that do NOT count (ignore them): * Speed-limit signs * Street-name signs * Parking-rate / hours signs * Weight limits ("GVW 30,000 LBS") * Billboards, storefront signs If no clearance sign is visible in any image, set found_sign=false. Do NOT guess based on the structure type.

This matters because an LLM that's been told "tell me what you see" will cheerfully make something up. Our prompt instructs it to answer "no sign" when the sign isn't there, and to flag partial views as low confidence.

4

Only write if the model is certain

By default we only commit a reading when confidence === "high". Anything medium or low is logged for a human to spot-check later. The script has a --confidence medium flag we can use on a case-by-case basis, but the production pass is high-only.

Before writing, two hard sanity bounds:

  • Readings outside 4 ft – 20 ft are treated as parse errors and discarded.
  • If the new reading differs from our existing (non-AI) value by more than 2 ft, it's held for manual review rather than overwritten.
5

Record provenance, not just the number

Every AI-verified entry keeps:

  • source: "AI-verified (Street View + Claude Vision)"
  • verified_on: "2026-04-18" — ISO date so the app can stale-check later
  • A note appended to the entry with the exact sign text Claude read

The AI-verified badge you see in the app is driven by that source field. No behind-the-scenes relabeling.

What this is not

This is not a substitute for looking at the sign. AI vision is good — on our test set it agrees with a human reader about 90% of the time — but 10% is still wrong 1 in 10. Wrong about a clearance, at the wrong moment, costs you a torn-off AC unit, a shattered windshield, or worse.

Every time you drive toward a garage, the posted sign at the entrance is the only authoritative number. Our data — AI-verified or otherwise — is for planning. The sign is for committing.

Why we think it's still worth doing

Before AI verification, about 38% of the garages in our database had no posted clearance from any upstream source (OpenStreetMap, operator websites, FHWA). Those entries rendered as "Unverified" — which is honest, but not useful.

AI verification converts most of those into real numbers you can filter and sort on. The remaining 10% of AI errors get surfaced two ways:

  1. User reports. Every garage has a "Report clearance" button. When a user tells us our number is wrong, the AI-written value gets flagged for immediate re-verification.
  2. Periodic re-runs. The verified_on date drives a "data may be stale" banner after two years, at which point we re-fetch Street View (which may itself be fresher imagery) and re-ask.

What data we keep about you

None. The AI-verification pipeline is entirely offline — it runs against Google Street View and Claude, not against your session. We don't send your location, vehicle height, or any identifying information to the AI. See our privacy policy for the full disclosure.

Technical details, if you're curious

Questions or corrections

Spot a bad AI-verification? Open the entry in the app and click Report clearance. It goes to a queue we review daily.