Your data stays yours.
Most of what Parkway does needs no data from you at all. When you connect your own systems, the analyst sees only aggregate findings, and you can run the whole thing inside your own environment.
Less of your data than you would expect.
Nothing from you
Cross-park intelligence runs on the Thoosie network: posted waits, ride status, weather forecasts, ride specs and park maps. Your systems are never touched.
Aggregates only
Connected sources are reduced to aggregate findings before the analyst sees them. Deploy inside your environment and raw records never leave it.
Scope you control
Each source is connected with your approval, at the scope you choose, and can be disconnected.
Hosted by us, or inside your cloud.
The language model only ever sees the results of the analyses, never your raw data. Where it runs is your choice.
- Our hosted service, for the quickest start.
- Your own cloud account through Amazon Bedrock or Google Vertex AI, so nothing reaches a third party.
- A local model on your own hardware.
- Anthropic does not train its models on commercial customer data.
Access follows your organization.
Your park, your team
Sign-in by park, with permissions set per person. Settings and saved peer groups belong to your park.
Your parks only
A chain sees its own parks side by side. It never sees another operator's parks.
For your own team
Ride-by-ride downtime detail is for the park's own operators, behind accounts and permissions you control.
The same care, pointed outward.
- Peer parks are counted, never named, and every group is large enough to keep its members anonymous.
- It does not track individual guests. Its crowd signals come from posted wait times and ride status.
- The screenshots on this site come from a real park, with the park's name and its rides' names hidden.
What security teams ask.
Do you need access to our systems to start?
No. Tier 1 runs entirely on cross-park data. Connecting your own systems is optional and comes later, at a scope you approve.
Does the language model see our raw records?
No. It sees the results of the analyses, which are aggregates, never raw records.
Can it run without sending anything to a third party?
Yes. The model can run in your own cloud account through Amazon Bedrock or Google Vertex AI, or as a local model.
Where are the exact terms?
Data handling, retention and security terms are confirmed in contracting. Ask us and we will send them for your review.
Bring your security questions.
We will walk your team through exactly what is connected, what is stored and where the model runs, before anything is turned on.