VoyageSafe

Our team is bringing a service, a website AI chatbot, that can provide travel suggestions based on your preferences while considering safety for everyone, especially women who want to travel alone.

Comment

Inspiration

We got the idea from our communities. Many women have talked about how they feel better traveling alone but have to find at least one travel partner for safety concerns, which can limit their travel experience. There is always intimidating news about traveling alone. However, traveling should be encouraging, educational, and even empowering for women, which makes us want to build something that not only makes women feel safe when traveling but also encourages a more friendly traveling service environment.

What it does

Our product is called VoyageSafe, an AI chatbot that helps generate travel routines for each customer. It would start by asking for the destination of the trip and the travel preference. Then, there will be more questions to follow up to narrow down the search range. The goal is to provide travel routines that value safety and individualized design.

How we built it

Our development journey began with creating a prototype in Figma, marking a new experience for our team. This initial design phase helped us visualize and refine the user interface before moving into active development. For our AI implementation, we leveraged Google AI Studio to build our conversational system, choosing Gemini 1.5 Pro as our foundation model. To ensure user safety and comfort, particularly for our target audience of women travelers, we implemented strict content filtering by selecting the "block most" safety setting. We configured the model with a temperature setting of 0.7, striking an optimal balance between creative and focused responses. This configuration allows the model to generate varied yet relevant responses while maintaining conversational consistency and natural dialogue flow. The user experience begins with specifying their desired travel experience, which helps our system contextualize and personalize subsequent recommendations. This thoughtful approach ensures that all advice is tailored to each user's specific needs and preferences. To provide comprehensive travel information, we integrated several key APIs. The Google Maps API forms the foundation of our location services, providing detailed information about points of interest and neighborhood assessments. For accommodation data, we implemented a dual-source approach using both Booking.com's API through RapidAPI and the Amadeus Hotel API. This combination gives our users access to extensive hotel options and crucial safety information, with Booking.com providing verified guest reviews and safety ratings, while Amadeus offers access to global hotel inventory, competitive pricing, and detailed property information. By cross-referencing these sources, we can provide users with the most comprehensive and reliable accommodation data available. The technical implementation was built using Next.js as our framework, with JavaScript powering our web interface to connect the various APIs seamlessly. Our development process followed a systematic approach, starting from the initial Figma prototype, moving through AI model implementation and API integration, and finally to the web interface development. Throughout the development process, we conducted multiple rounds of testing and made iterative improvements to the UI based on user feedback. These refinements helped us optimize the user experience and ensure that our platform effectively serves its purpose of helping women plan safer travels. After completing all the coding, Sara pushed all the code to GitHub. Once everything was uploaded to GitHub, we proceeded with deployment on Vercel. I purchased a domain name (https://voyagesafe.click/) from NameSilo.com. To manage the domain's DNS settings, I used Cloudflare through dash.cloudflare.com. Finally, I returned to Vercel to add my custom domain name to the project, successfully connecting the domain to our deployed application.

Challenges we ran into

Time Our team faced several significant challenges while developing our travel safety platform, with the primary focus on delivering quality work under tight time constraints. We are both college students, so it's about how to find an intelligent balance in how we distribute our time for our tasks. Although this is our first chance to learn about coding and other related technologies, we still expect a certain quality of our work. That sometimes makes it trickier for us to manage our time because the time on a project increases when it's more technical.

Technical issues Learning and implementing new technologies presented a substantial learning curve. Our first experience with Figma for UI/UX design required significant investment in understanding the platform's capabilities and best practices. While this initially impacted our development timeline, it ultimately proved valuable for visualizing and iterating our ideas effectively. Our most significant technical challenge centered on integrating three distinct APIs - Google Maps, Booking.com, and Amadeus Hotel API - each with unique documentation, authentication methods, and data structures. The team faced various technical difficulties, particularly in coordinating these data sources into a cohesive system while managing API rate limits, authentication security, and data normalization. Beyond the API integration, we struggled with translating our conceptual ideas into functional code. As the complexity of our system grew with additional API integrations and features, ensuring smooth functionality became increasingly challenging. However, through persistent troubleshooting and methodological adjustments, we successfully overcame these obstacles and achieved a result that met our quality standards.

Accomplishments that we're proud of

We are proud of every steady step we made in this project, especially since it's a totally new field for our team; there's a great sense of achievement when we are actually building something that we believe can help people in some ways and we are bringing the idea to life(to a certain extent). The first time doing UI design, logo design, coding, etc., it was a challenging but incredibly remarkable experience for our team. We are also delighted and proud of the final result, which achieved the initial planned functions. And the minimalist style doesn't look ugly at all!

We are also very proud of how supportive our team members are to each other along the process. The cooperation and the communication were very efficient and considerate. And we also divide the work according to our interests and specialties. Still, our team members are always ready to stand up with a sense of responsibility when another teammate needs it.

What we learned

Some of the significant learning outcomes for us are:

  1. The increasing knowledge of how to communicate and use all those handy and intelligent tools wisely, along with the newest developed tools like Google Gemini and other design tools.
  2. Develop a sense of logic regarding how the start of the project and developing a prototype should be, and learn to start making a reasonable to-do list before actually getting to work.
  3. Realize a reasonable schedule for to-do lists, and always the sooner, the better, so we save time for unexpected things during the project.
  4. Never easily give up and embrace technical difficulties because we always learn from those and are potentially aware of how to avoid them in the future.
  5. A supportive and responsible team is so important for the progress and well-being of the project. Knowing how to care for teammates will make the work more efficient than expected.

What's next for VoyageSafe

We plan to work more on the technical aspect of the VoyageSafe chatbot and train it to be smarter. We are also trying to make the page more visually pleasant while developing new functions, such as booking hotels and tickets through our chatbot, to make it even more convenient for our customers.

Share this project:

Updates