AccessPath checks every critical part of a journey against attributable evidence, exposes uncertainty, and replans around barriers.
01UnderstandYour access needs
02VerifyOfficial evidence
03ReplanAround barriers
Plan a verified journey
What do you need?
This pilot evaluates one evidence-backed London journey and lets you change critical requirements.
Initial route loaded with 25% confidence. Verification is ready to run.
Accessibility assessment
Initial route found. Verification pending.
Curated evidence snapshot
Pre-verification finding
The initial route contains conflicting entrance evidence
Run verification to apply your requirements, inspect the conflict, and reveal any usable alternative.
Initial route
Victoria to the Natural History Museum
5-hour visit
Select any step to inspect its claims, sources, freshness, and score.
Trust through transparency
AI reads meaning. Rules decide safety.
AccessPath separates semantic reasoning from deterministic accessibility decisions.
01
Structure the request
A schema-constrained AI provider extracts needs such as step-free access and required facilities.
02
Connect the evidence
Claims are linked to the exact entrance, vehicle, venue, or facility they describe.
03
Score deterministically
Source authority, directness, freshness, corroboration, and conflicts determine confidence.
Reliable fallback activeThe initial demonstration uses validated default constraints. Submit the form to run the configured AI provider.
Before you travel
Accessibility conditions can change. This demo uses a curated evidence snapshot, not live operational status. Confirm critical details with TfL and the venue before travelling.