ParkPass AI
Just tell ParkPass what you need.
Staff set things up by asking. Managers get a written morning briefing that flags problems. Visitors get a booking assistant in their own language. All of it governed for compliance.
AI you can put in front of an auditor
Every assistant works inside the permissions, audit trails and payment rules your park already relies on.
Works through your permissions
ParkPass's assistants work through the same secure ParkPass APIs and permissions as your staff. They have no direct access to your database.
Acts as the signed-in staff member
An assistant can only do what that person is already allowed to do, and its access ends when they sign out. It starts read-only until your park switches changes on, and ParkPass's servers enforce that, not the AI.
Guest data off-limits by default
Guest personal information is off-limits to the AI by default, even when the staff member can see it. Staff accounts, passwords, sign-ins and permissions are always off-limits.
Never sees card details
Assistants never see card numbers. Visitors always pay on your own secure checkout, which is what keeps the AI out of the payment path entirely.
Reports built from summary data
Guest names, emails, phone numbers and addresses are never sent to the AI, and staff are never named in reports. Your data is never used to train AI models.
Recorded, reviewable, stoppable
Every change is recorded under the staff member's name, and every conversation is saved. Each park has an on and off switch that stops the AI instantly, plus usage limits.
Your choice of AI model
ParkPass AI is model-agnostic. Run it on leading models from Anthropic, OpenAI or Google, on a model hosted in your own country, or on a private model running on your own infrastructure so data never leaves it.
ParkPass is PCI DSS and SOC 2 compliant and independently penetration-tested every year. Card data never touches ParkPass, and your data stays in the region you choose.
Every ParkPass AI feature runs on Orcheza, ParkPass's AI agent platform, built to keep AI accountable.
Read how Orcheza governs AI agentsSet up by asking
Skip the training manual. Just ask.
The ParkPass Assistant is built into the admin dashboard. Open it from any screen and describe what you need.
- "Where do I refund a deposit?" gets the exact menu path and a button that opens that screen, anywhere in the dashboard.
- Describe a change and it makes it: destinations, open and close dates, time slots and capacity, prices, reservation types, visitor types, products and add-ons, booking forms, coupons, memberships and gate schedules.
- It shows exactly what it will change and asks you to confirm before anything happens.
- Big jobs, like updating prices across every location, run in the background and report back. Changed your mind? Ask it to put things back.
- New and seasonal staff are useful on day one. Nobody has to learn every screen.
One request
Know what happened yesterday before your first coffee.
Every morning an AI analyst reviews yesterday's bookings, visits, memberships, cancellations and refunds, gate activity, weather and staff activity, and writes a plain-language briefing your team reads in ParkPass. A weekly business review is emailed to managers.
ParkPass calculates every number from your records; the AI explains them. Every item it flags links to the exact reservations or tills behind it, so staff can check.
Every morning
Your morning briefing is ready
Yesterday's bookings, refunds, tills and gate activity, in plain language, with the items worth a look on top.
Ask follow-up questions
- Ask in plain words: "How did this long weekend compare to last year?"
- Demand forecasts combine your history, holidays and the weather forecast.
"Worth a look" alerts
- A spike in refunds or cancellations, or unusual discount and coupon use.
- A till that did not balance, or cash left in a drawer.
- A gate that dropped offline, or plate reads failing.
AI fraud patrol
- An AI agent runs on a schedule, looking through your data for abuse.
- Shared memberships and passes: the same pass scanned at two gates minutes apart, or used far more than one household could.
- Promo codes leaking online, refund and chargeback patterns, duplicate or throwaway accounts, and resale.
- It flags each case with the evidence behind it and, under your park's rules, blocks it: it suspends the pass or account and records why.
Automated safeguards
AI finds it, automated safeguards stop it:
- Bot screening at online checkout, and bank verification (3-D Secure) with stricter checks for high-value orders.
- A web application firewall that blocks known bad traffic and slows attacks on sign-in.
- Two-step sign-in for staff, automatic shut-off of unused staff accounts, and per-ticket check-in limits.
- Gates that read licence plates from camera images and admit only valid, unused bookings, plus membership rules that stop pass sharing and double booking.
- Card payments are reconciled against bookings every day, and every night ParkPass checks your payments against your revenue records.
On your booking site
A booking assistant that knows your rules.
A chat on your booking site, open 24/7, helps visitors find open dates and the right ticket, pass or program, answers questions from your own prices and policies, and builds the booking for them, holding it while they review.
- It follows the same capacity, pricing, visitor-type and member rules as your booking page, because it uses the same system.
- Visitors always review and pay on your own secure checkout; the assistant never touches card details.
- For a change of plans, it takes them to their secure booking link, under your change policy.
- Signed-in families can ask what fits their kids and have the cart filled for them.
- For the front desk too: staff can ask the assistant to look up a family, sell a pass or pause a membership, with a confirm step before anything is charged.
Any language your visitors speak.
- The assistants understand and answer in whatever language someone writes in, from French and Spanish to Punjabi, Mandarin, Tagalog, Arabic or Ukrainian.
- There are no translation files to maintain and nothing to set up. Visitors can switch language mid-conversation.
- Staff can ask the admin assistant in their own language too, and the morning briefing is written in the language your team works in.
- Booking pages already come in English, French and Spanish; the AI goes beyond that to whatever language the visitor uses.
No setup needed
The engine underneath: Orcheza.
Orcheza is ParkPass's AI agent platform, and every ParkPass AI feature runs on it. It is how each assistant stays accountable.
Sandboxed by design
Every agent runs sandboxed in its own container, with only the tools and credentials its job needs.
Tied to one person
Agents work in ParkPass through short-lived access tied to one staff member, and human approval is a built-in workflow step, not an afterthought.
Switches, limits and logs
A per-park on and off switch, spending limits and cost tracking. Changes and approvals are recorded in an audit log, and the control dashboard is reachable only over a private network.
Model-agnostic by design
Swap AI models without changing anything else, including a private model running on your own infrastructure.
Works where your team works
The same assistants can meet your team where it already works: Microsoft Teams, Slack or Google Chat, as well as ParkPass.
We run ParkPass on it
Orcheza agents write, review and test ParkPass code, prepare every release for a person's approval, draft our RFP responses, and maintain this website. About three quarters of the changes to parkpass.ca since July were made by Orcheza agents, each one approved by a person before it went live.
See ParkPass AI on your own data.
Bring a season's worth of questions and we'll answer them live in a personalized demo.