Inclusive Travel Agent

Travel excites, but disabled travelers face many barriers. The AI Inclusive Travel Agent eases this with personalized, verified accessible itineraries in minutes, covering flights, hotels.

Comment

Inspiration

Travel is meant to be freeing — yet for millions of disabled travelers, it is stressful, inaccessible, and often exclusionary. From inaccessible websites, unclear mobility information, to overwhelming trip planning, many disabled people avoid traveling altogether.

Our team was inspired by one simple question: “What if an AI agent could plan a fully accessible and inclusive trip in minutes — tailored to each person’s disability needs?”

This project was born from lived experiences, stories of disabled travelers, and a shared mission to make travel inclusive, barrier-free, and joyful for everyone.

What it does

Inclusive Travel Agent is an AI-powered multi-agent travel planner that designs customized, accessibility-verified itineraries for disabled travelers.

It serves as a smart accessibility-aware travel companion that:

Understands a traveler’s accessibility needs (e.g., blind, deaf, wheelchair user, neurodivergent, etc.)

Researches destinations, transportation, hotels, attractions, and services with accessibility criteria

Verifies accessibility information (e.g., ramps, elevators, tactile paving, sign language support)

Suggests best routes and mobility options (walking, bus, taxi, paratransit) with cost comparison

Generates a ready-to-use inclusive travel itinerary within minutes

The result: a personalized, inclusive, and stress-free travel plan, without the hours of searching, uncertainty, or risk of inaccessibility.

How we built it

We built the platform in three iterative phases, prioritizing user feedback and technical scalability: User-Centric Design: We interviewed 23 individuals with diverse disabilities (mobility, visual, auditory, neurodiverse) to map pain points, which informed our agent architecture. This led to a modular system where a root_agent orchestrates 6 specialized agent categories (inspiration, planning, booking, pre-trip, in-trip, post-trip). Technical Stack: Frontend: A React-based interface with accessibility features (screen-reader compatibility, adjustable text size). Backend: FastAPI for APIs, with agent logic built using LangChain for flexible prompt engineering and task delegation. Integrations: Connections to external databases (e.g., Wheelmap.org for accessibility data), airline/hotel APIs, and communication tools (email/SMS for real-time alerts). Architecture: A layered system separating user interfaces, core logic, automation agents, and external integrations (see system architecture diagram) for easy updates. Testing & Refinement: Beta testing with 10 users helped refine critical features—for example, enhancing the in_trip_agent to handle last-minute accessibility emergencies (e.g., broken elevators) by cross-referencing nearby alternatives.

Challenges we ran into

Fragmented Accessibility Data: Many travel providers lack standardized APIs for accessibility information. We addressed this by combining ethical web scraping, partnerships with disability advocacy groups, and user-generated content moderation. Agent Coordination: Ensuring 15+ specialized agents (e.g., flight_seat_selection_agent and payment_agent) communicated seamlessly required building a shared context log managed by the root_agent. Edge Cases: Unexpected scenarios (e.g., a hotel misadvertising accessible rooms) demanded training the AI on 500+ crowd-sourced “travel emergencies” to improve response accuracy. Trust in Automation: Users were initially skeptical of AI-generated accessibility claims. We mitigated this by adding “verification badges” for provider-confirmed details and a “feedback loop” for users to flag inaccuracies.

Accomplishments that we're proud of

Built a working prototype of an AI Multi-Agent system tailored for inclusive travel planning

Created a first-of-its-kind concept focused on disability-inclusive travel using AI

Demonstrated how AI can remove barriers and empower disabled travelers

Designed a workflow that is scalable to global regions and multiple disability types

What we learned

How multi-agent AI systems can solve complex real-world inclusion problems

The importance of human-centered design when building for disabled users

Accessibility is not “one-size-fits-all”—personalization is key

AI can significantly reduce cognitive load and planning time for disabled travelers

What's next for Inclusive Travel Agent

We’re excited to grow this project beyond the hackathon and evolve it into a powerful, everyday companion for disabled travelers. Our next steps include:

Integrating with existing navigation ecosystems (e.g., Google Maps, Apple Maps) to enhance accessibility layers and provide step-by-step inclusive navigation, especially for blind and visually impaired users

Building a personalized AI navigation assistant that adapts to different disabilities and individual lived experiences — offering tailored mobility guidance, anxiety-free trip support, and real-time problem-solving on the go

Expanding from trip planning into a holistic travel-life assistant that supports disabled users across the full journey: planning, traveling, navigating, enjoying, and staying safe

Co-designing with global disability communities to ensure cultural, regional, and disability-specific inclusivity

Developing a scalable ecosystem that works across countries, transport systems, and accessibility infrastructures

Our long-term vision is bold but simple: Make travel easy, accessible, and fun for 1.3 billion disabled people — anywhere in the world.

Built With

Share this project:

Updates