An AI-enhanced Incident Management System that tells your control room what's urgent, what's a pattern, and what can wait — in plain English, with the reasoning shown.
Pilot and production deployments run in UK AWS infrastructure (London, eu-west-2).
Certified. Per-organisation data isolation and role-based access as standard.
AI surfaces, scores and explains. Your team acts. No automated changes to your systems.
PathFinder AI is a member of the NVIDIA Inception program, supporting our product and go-to-market development as we build explainable incident intelligence for UK operations teams.
Membership in the NVIDIA Inception program does not imply endorsement by NVIDIA.
For social announcements, we use #NVIDIAInception.
NVIDIA and the NVIDIA logo are trademarks and/or registered trademarks of NVIDIA Corporation in the U.S. and other countries.
Familiar in any busy control room
Radio, email, telematics, monitoring dashboards, phone calls. The full picture lives in six places at once.
When every alert is "critical", none of them really are. Controllers end up ranking by whoever shouted last.
Each incident is handled and closed. The pattern underneath it stays invisible until it happens again.
The experienced controller knows which incidents matter. That judgement is rarely written down anywhere.
PathFinder doesn't replace your controllers.
It gives them clarity.
Turn incident noise into a clear, ranked operational picture
PathFinder sits above your existing systems — no rip-and-replace, no automated actions.
Webhooks, email, APIs, existing monitoring and alerting feeds. One connection per source, one unified view.
PathFinder assesses impact, history and pattern. Every incident is scored 1–10 with its reasoning attached.
See what's urgent, what's a pattern, what can wait. PathFinder recommends — your team chooses and acts.
“PathFinder suggests, surfaces, and explains — humans decide.”
Illustrative examples of how PathFinder ranks a live operational picture, with the reasoning shown
Blocks a stand at peak loading. Three connecting services affected and passenger build-up already reported at the stop.
Real-time position data incomplete, which degrades passenger information and every downstream decision made from it.
Leading indicator of cancellations at first departure. Matches the pattern seen on the same turn in the previous fortnight.
Localised service quality impact with no effect on vehicle movements. Alternative information channels remain available.
Advance notice with adequate lead time. No action required today; flagged for scheduled diversion planning.
Every score carries its full reasoning and is logged for audit. No black boxes.
Everything needed to turn incident noise into an actionable operational picture
Every incident assessed and ranked 1–10 on impact, history and pattern, so the queue reflects what actually matters.
Suggested next steps with the context behind them. Controllers decide whether to act — PathFinder never acts alone.
Surfaces the recurring problems behind individual incidents, turning reactive handling into evidence for change.
Webhooks, email, APIs and existing monitoring feeds brought into a single prioritised view of the operation.
Every score shows why it was given and is logged with its reasoning — reviewable after the event, not a black box.
Data isolated per organisation with role-based access, hosted on UK infrastructure.
Not just technical incidents.
PathFinder handles operational incidents too — service disruption, vehicle and asset faults, staffing shortfalls, and safety-related reports.
Designed around control room workflows from the start.
Purpose-built for control room workflows, unlike generic incident management tools. Alerting platforms tell you something happened; PathFinder tells you what it means and how it ranks against everything else that is live.
No automation surprises. AI recommends, controllers decide.
PathFinder takes no automated action on your systems or your fleet. It gives recommendations with the full context attached, and your team chooses what to do with them.
UK company. Cyber Essentials certified.
Built in the UK, with pilot and production deployments made in UK AWS infrastructure (London, eu-west-2) to meet public-sector expectations on data residency and governance.
Every score explained and logged with its reasoning.
When a decision is questioned after the event, the reasoning behind the priority it was given is on the record rather than reconstructed from memory.
What procurement and governance teams need to know
Built around bus control centre workflows, not adapted from generic IT service-desk software.
Pilot and production deployments are made in UK AWS infrastructure (London, eu-west-2), meeting public-sector data governance requirements.
With per-organisation data isolation and role-based access control as standard.
PathFinder surfaces, scores and explains. Controllers decide. No automated action is taken on your systems.
PathFinder is in active pilot with transport operators — .
Cloud-native, multi-tenant architecture
Data isolated per organisation
UK AWS infrastructure for pilot and production (London, eu-west-2)
See PathFinder scoring live incidents against your own operational context. We'll walk you through the platform and scope a pilot with you.