Inspiration
With an overwhelming amount of information available online, travelers often struggle to decide on the best activities, attractions, and destinations to visit. With JourneyGenie, travelers can save time and reduce the stress of trip planning, while also ensuring that they make the most of their time and have a memorable experience. By incorporating both practical recommendations and creative visuals, we strive to create an unforgettable travel planning experience that caters to the diverse needs and preferences of every traveler.
What it does
JourneyGenie solves these problems by providing personalized recommendations for each day of the trip based on the traveler's preferences and interests. The data collected from user's response is analyzed using natural language processing algorithms powered by Cohere to create customized recommendations that include the best activities and attractions. We also provide artistic images of the locations, designed to help travelers visualize the destinations in a way that resonates with their unique preferences and interests.
How we built it
To create an intuitive and visually appealing interface, we utilized Figma for prototyping and React for the frontend to allow for interactive user inputs, resulting in a seamless and streamlined experience for travelers.
Furthermore, we utilized Cohere's API to help us generate 5 personalized recommendations by analyzing the user's travel personality. Our team carefully selected seven questions to accurately determine the user's preferences, ensuring that the recommendations we provide are both relevant and useful.
To create a truly immersive and inspiring experience, we also utilized Open AI's Dall-E technology to generate artistic images for each destination. By displaying these images when the user selects a city on the map, we help them to visualize their upcoming trip and to get excited about the possibilities. Finally, we send a summary of the places to the user's mobile phone through a text message, using Twilio.
Challenges we ran into
Generating accurate recommendations using Cohere was a major issue. Deciding the questions which accurately represent the personality of the traveler helped us improve.
Accomplishments that we're proud of
We are glad to have this working model which can suggest the best places to visit in over 125 cities in the world based on the traveller's personality. The model can also save the personality type of user and then user can keep selecting different locations for recommendations. The tested runs had image outputs which were accurate depiction of cities.
What we learned
We learned about travel personalities and prompting the LLMs with which precise instructions would best decipher determine travel personality. We also learned how to integrate Cohere and Dall E API into interactive user experience.
What's next for JourneyGenie
We look forward to display review of previous visitors and create an image of their experience. This would help users also visualize a realistic experience they may expect for each location
Built With
- cohere
- dall-e
- domain
- domain.com
- figma
- javascript
- react
- twilio
- typescript
Log in or sign up for Devpost to join the conversation.