Use Case

Virtual travel itinerary planning

Generative AI models like Language models (LLMs) and image generators can revolutionize virtual travel itinerary planning. For example, an AI tool like ChatGPT could help users articulate their travel preferences, and by training on rich travel data, could generate tailor-made travel itineraries. Users can specify their desired destinations, durations, activities or budget, and the AI can create a detailed plan including immersive descriptions of destinations, key attractions, optimal routes and recommended eateries.

Image generators can enliven itinerary planning as well. They can curate a visual walkthrough of the proposed route or experiences, generating realistic renditions of tourist sites, accommodation interiors or authentic local cuisines. This visual narrative could provide users a virtual ‘try-before-you-buy’ experience, vastly improving the satisfaction and success of travel planning.

How to build with Clevis

Text Input

Ask the user for their travel destination

Prompt ChatGPT

Prompt ChatGPT to provide recommendations for popular attractions and landmarks in the destination

HTTP Request

Fetch images of the recommended attractions and landmarks

Generate Image (DALL-E)

Generate a collage of the fetched images to display in the itinerary

Display Output

Display the generated itinerary with the recommended attractions and landmarks along with the generated image collage

This example application, created using Clevis, is known as the Virtual Travel Itinerary Planner. Much like the name suggests, it aids users in planning a virtual itinerary for their chosen destination. Here is a brief explanation on how it works:

Firstly, it accepts text input from the users about their preferred travel destination. Post this, it prompts ChatGPT to generate recommendations for popular tourist attractions at the specified location. Using this output from ChatGPT, it proceeds to make an HTTP request to fetch images for these recommended attractions.

In the default configuration, the request is sent to 'https://api.example.com/images' via a GET method, and includes authorization headers along with query parameters carrying the suggestions from ChatGPT. Following the request, the fetched images are used by DALL-E, a subsystem within the AI application, to create a collage.

As a final step, this developed itinerary carrying the recommended attractions, alongside the generated collage, is displayed to the user. This is an example of the multitude of applications developed within the same area via Clevis, relying on the robustness of ChatGPT and OpenAI.