Skip to main content

Trip Planner Example

Trip planning is a useful showcase for multi-step AI because it naturally combines research, data retrieval, and itinerary synthesis. It is also a good way to understand the boundary between what community can validate today and what becomes stronger with enterprise-only connectors or protected operations workflows.

Community-Friendly Architecture​

A public/community travel assistant can still demonstrate the main AxonFlow ideas:

  • generate and execute a MAP plan
  • call governed MCP connectors for travel-adjacent data sources you already control
  • synthesize itinerary output through governed LLM requests
  • record the execution history for debugging and audit review

The most realistic community approach is to pair MAP with connectors such as PostgreSQL, HTTP, or Redis, then reserve enterprise-only travel connectors for later licensed rollout.

Why This Is a Useful Trial Scenario​

Trip planning is not just a travel demo. It stresses several important AxonFlow concepts at once:

  • multi-step decomposition
  • governed retrieval from several data sources
  • synthesis of a user-facing result from many intermediate steps
  • execution visibility across a workflow rather than a single request

That is why teams often use a trip planner or research planner as a trial workflow even when their actual production system is in another industry.

Example: Generate and Execute a Travel Plan​

import { AxonFlow } from '@axonflow/sdk';

const client = new AxonFlow({
  endpoint: process.env.AXONFLOW_ENDPOINT!,
  clientId: process.env.AXONFLOW_CLIENT_ID!,
  clientSecret: process.env.AXONFLOW_CLIENT_SECRET,
});

const plan = await client.generatePlan(
  'Plan a 3-day trip to Amsterdam with hotel, restaurant, and activity suggestions',
  'travel'
);

console.log(plan.planId);
console.log(plan.steps);

const execution = await client.executePlan(plan.planId);
console.log(execution.status);
console.log(execution.result);

Why This Example Matters​

This example helps engineers answer two high-value questions:

  • can AxonFlow coordinate a multi-step planning workflow end to end?
  • can we keep governance and execution visibility while the workflow grows more sophisticated?

That makes it a strong trial scenario for teams building internal assistants, itinerary tools, or more general research-and-planning systems.

Community Path vs Licensed Path​

The community path is strong enough to validate MAP itself, the plan lifecycle, and the way governed data retrieval fits into a multi-agent workflow. The licensed path becomes more relevant when the rollout needs travel-specific connectors, richer approval workflows, or broader operations tooling for teams beyond the core builders.

When Evaluation or Enterprise Helps​

Evaluation or enterprise becomes much more compelling when the travel assistant needs:

  • enterprise-only travel connectors
  • approval workflows for higher-risk bookings or budget thresholds
  • protected portal workflows for operations and review
  • deeper governance and observability across many tenants or business units
Last updated on