Hosted workspace
Your team works in Nova's hosted workspace: proposals, routing, booking queues, signals and conversations, with single sign-on, roles and an audit trail. Nothing to install.
Platform
One AI travel platform, built on what's actually possible
Six modules, from the first proposal to in-trip re-planning to checkout by AI agents, share one data layer that knows your opening windows, capacity, policies and pacing. Start with the module that hurts most. Add the next when it earns its place.
The platform
Six modules, one living plan
Each module works on its own. Together they turn an itinerary from a document into a plan that can be routed, booked, predicted and re-planned.
Operational truthOne shared model of suppliers, inventory, travelers, places, time and weather
- Itinerary StudioReads a traveler's half-said wishes and drafts a routed, bookable multi-day proposal from your own product library. Your advisor edits, approves and sends it.
- Route EngineSequences every day against opening windows, travel times by mode, pacing budgets, accessibility needs, weather exposure and supplier capacity, and explains each choice.
- Booking FlowSends supplier requests, reads confirmations in whatever format they arrive, issues vouchers and reconciles payments. Anything odd is lifted aside for a person.
- SignalsForecasts demand, flags cancellation risk and price sensitivity, and finds the demand hiding in your shoulder season. Every signal comes with a suggested next step.
- ConciergeAn in-trip assistant on web chat, WhatsApp, SMS and email that answers from your live data, re-plans when the day breaks, and knows when to hand off to staff.
- Agent GatewayYour own MCP server and agent-safe checkout, so AI assistants can find and book you directly under your policies, spend limits and approval rules.
A plan, not a PDF
Most travel software treats the itinerary as a document. It gets written once, sent, and patched by phone when the trailhead closes. We treat it as a living plan: something that can be routed, predicted and re-planned, because every module reads the same facts.
That’s the whole design. When Booking Flow learns a supplier has scrubbed tomorrow’s 6:40 AM balloon launch, Route Engine already knows what else fits that morning. Concierge can explain the change to the traveler in plain words, and Agent Gateway stops offering the slot to AI assistants. Nobody re-keys anything, and there is one version of the truth.
At every point where a decision matters, whether that’s money, safety or a change to someone’s holiday, the platform drafts and a person decides.
Shared data layer
Operational truth, laid down in layers
Underneath the modules sits a single model of what is actually possible on a given day, for a given traveler. We build it with you, layer by layer, from the systems you already run. Every module reads from it and writes back to it, so a fact learned once is known everywhere.
- Supplier truth. Products, contracts, allotments, cut-off times and cancellation terms, deduplicated across feeds.
- Place truth. Opening windows, seasonal closures, accessibility details and weather exposure, verified by people who have been there.
- Movement truth. Travel times by mode and hour of day, transfer buffers and lower-carbon alternatives.
- Traveler truth. Consented preferences, pacing and access needs, kept to the minimum and access-controlled.
- Live truth. Weather, closures, delays and supplier changes, timestamped so every answer knows how fresh it is.
How modules combine
Three combinations we see most often
Each is a starting shape, not a bundle. Which module comes first depends on where your hours and your risk actually sit.
- Itinerary Studio · Route Engine · Booking FlowThe proposal-to-confirmation deskFor operators, DMCs and advisors. Studio drafts from your library, Route Engine checks that every day holds together, and Booking Flow sends supplier requests the moment the traveler says yes. The usual first win is proposal time.
- Route Engine · Concierge · Booking FlowThe adaptive tripFor outdoor and weather-exposed itineraries. A storm cell is forecast over the high trails: Route Engine re-sequences the afternoon, Concierge proposes the change to the guide and the traveler, and Booking Flow tells the two affected suppliers.
- Agent Gateway · Signals · Itinerary StudioThe agent-ready destinationFor destination organizations and lodges. AI assistants query your catalog through Agent Gateway, Signals steers them toward quieter dates and places, and Studio packages what's worth recommending. The usual first win is being findable at all.
Deployment options
Run it where your team and your travelers already are
Most customers combine two of these. The data layer underneath is the same either way.
Embedded widgets
A trip planner, availability view or Concierge chat embedded in your own website and booking pages, styled to your brand and reporting into your analytics.
REST API and webhooks
Call Route Engine, Signals or Booking Flow from your own systems and subscribe to events when plans change. Versioned and documented, with sandbox access on request through Developers.
MCP server
Agent Gateway exposes your catalog, policies and checkout to AI assistants over the Model Context Protocol, the open protocol for connecting assistants to tools and data.
How it works
How teams usually start
Every engagement is scoped and quoted after a discovery call. The full picture is on engagement models.
- 01
Discovery Sprint
3–4 weeksWe map your operation, your data and your stack, then pick the module where the evidence says the first win is.
- 02
Pilot
8–12 weeksOne or two modules, live with real trips, measured against your own baseline. You see the numbers before you commit to more.
- 03
Platform subscription
AnnualThe modules that proved themselves, on a modular annual subscription. The data layer grows as you add modules.
- 04
Managed AI operations
OngoingWe evaluate, monitor and tune the AI in production through Managed AI Operations, so quality holds long after launch.
Questions
Platform questions
Do we need all six modules?
No. Most customers start with one or two. Each module works on its own, but because they share the data layer, every module you add starts from what the others already know.
Does Nova replace our reservation system or PMS?
No. Your systems of record stay in place. Nova reads from them and writes back. See integrations for the categories we connect and the patterns we use.
Where does the AI make decisions?
In as few places as possible. Modules draft, check, predict and propose. People approve anything that touches money above your limits, safety, or a change to a traveler's trip. That principle runs through our approach.
How do you handle traveler data?
Carefully, and as little of it as we can: least-privilege access, encryption in transit and at rest, minimization of sensitive fields, and no pooling of your data with anyone else's. Details are on security & trust.
Related
Keep exploring
- PlatformIntegrationsHow Nova connects to PMS, reservation systems, channel managers, payments, CRMs and messaging: REST, webhooks, CSV/SFTP and MCP, with written data contracts.
- PlatformSecurity & trustHow Nova protects traveler data and keeps AI grounded: encryption, least-privilege access, prompt-injection defenses, PII minimization and incident response.
- CompanyOur approachHow Nova Adventure implements travel AI: operational truth first, then routing, prediction and automation, with people approving every decision that matters.
- ResourcesDevelopersREST API, webhooks and MCP access to itineraries, routes, quotes, bookings and signals. OAuth 2.0, environment-scoped keys, sandbox access by request.
- WorkSelected workFour anonymized case studies from tour operators, destinations, lodges and event planners, with pilot results measured against a real baseline.
Start with the module that hurts most
Tell us where the hours go, or where trips break. We'll show you which module to pilot first, and which ones can wait.