AI BnB

• 

  Listings Dashboard.

• 

  Landing page, containing user's existing trip projects.

• 

  Chatbot interface for planning trips.

• 

  Module for listing generation using images of property.

Inspiration

As travelers, our group loathes planning process when it comes trips. Especially for trips that are longer in length, it takes a lot of time trying to source properties (Airbnbs) to stay in during the trip and being able to plan out exciting places to visit. We also sympathize with home renters, who have to engage in a lot of busy work in creating a listing just to make properties available. This being the case, we wanted to create AI BnB, which is an all in one AI utility that can be utilized by both travelers and home renters alike.

What it does

AI BnB streamlines the traveling process, allowing users to speak with a chatbot about what they are looking for in a trip (ex. destinations, number of people, budgets, activities, etc.) Taking all this information, our bot will create a traveling plan with an optimal set of AI BnB properties they can rent during the duration of their trip. We streamline the process for people to list their houses as well, where a listing can be generated after pictures of each room of the property are provided to our application. We also streamline collaborative trip planning as well, allowing people to invite each other to trips, and pitch in ideas for things to do on the trip, which our bot will combine together into a trip everyone can enjoy.

How we built it

For this application, we created these functionalities within a web application built on React + Next.js, with a Flask (Python) backend and Supabase database. In addition to these languages and frameworks, we utilized DedalusLabs very useful AI tools for streamlining usage of agentic AI and MCP servers.

Challenges we ran into

There were many struggles that came in creating this project, especially as a group of only two hackers. But nothing worthwhile comes easy. We had to spend the last two days battling with debugging environments and incompatible library versions. We had to use technologies we were never familiar with, such as agentic AI, to implement our product. And being able to balance such a big project amongst two people was challenging in being able to delegate UI, backend, database, and AI related tasks amongst ourselves in being able to create this bigger product.

Accomplishments that we're proud of

We are really proud that we were able to create something that is so practical and solves a real problem many face. In addition, it is a marvel that we were able to make a full-fledged application capable of creating these trips whilst having two people. The level of coordination we were able to achieve in being able to communicate, collaborate, work through adversity and ultimately achieve success was very satisfying, and we are very proud to have achieved what we did in 36 hours.

What we learned

When going into this hackathon, we wanted to create something different that wasn't done before. And in doing so, we had to learn how to be creative and try new things. We learned more about agentic ai and how useful it is in mediating and automating tasks. We learned about user design, taking feedback from many mentors at the hackathon. The biggest thing was learning to be okay with asking for help. We

What's next for AI BnB

We want this to be a full-fledged application that is accessible to as many people as possible, so it can make more happy and satisfied in their trip planning. That being the case, we want this application to come to mobile, where it can be available on people's iphone and android devices.

Built With

• dedaluslabs
• docker
• flask
• javascript
• next.js
• python
• react
• supabase
• tailwindcss
• typescript