Use Case #12 Personalized Travel using Personal Agents

[neiljthomson]

Personalized Travel using Personal Agents

Intro/Overview

The hospitality and travel industry is global in scope and is used by a large percentage of the population (who are affluent enough to do so). It is a prime example of an opportunity where personal data, requirements, and preferences could be leveraged by both personal (consumer-controlled) and service (service-provider-controlled) agents negotiating/matching personal wants and needs with service offerings.

Problems to Solve:

• Distinguishing and validating personal needs and wants provided directly by the traveler vs. inferred by indirect observations and assumptions about the traveler through their online behavior on travel and non or peripherally related sites and applications
• Integrating feedback (with agent assistance) on wants and needs from actual choices and behavior during past travel experiences into their current wants and needs
• Agent advice/agency for travelers on personal consent based on potentially complex requests for personal data across different jurisdictions and contexts.
• What is the threshold for allowing a personal travel agent to meet consumer expectations autonomously?
• Structuring of preferences, which may have complex context dependencies (e.g., type of accommodation or restaurant is dependent on type of trip (site-seeing vs. sports) and type of activities (winter/classic x-country ski)

Weaknesses in Current Practices:

• There are no established mechanisms for consumers to capture and present their personal data and wants and needs for a context like travel in a form that they can understand (and customize), which can also be consumed by LLM models consistent with the consumer's intent.
• Early-stage mechanisms (RAG) for incorporating human-readable/organized data (e.g., relational/graph) on personal data into LLMs such that LLMs do not hallucinate (partially due to large mutually conflicting wants and needs in the traveler's profile).

How Verifiable Data + Trust Spanning Protocol Can Help:

• Verifiable Data will be core to establishing trust between parties in the travel ecosystem (C2B, B2B), where a travel profile is a self-attested/signed credential, and verification between consumers and services will require extensive use of verifiable credentials.

What Is Not in Scope:

Additional Considerations:

Related Use Cases

• This use case builds on the Agentic AI use case, specifically on personalization and data privacy/user-controlled selective disclosure.