VentureVisionary

Inspiration

We were inspired to create something more than just a chatbot – we aimed for an interactive application that provided users with valuable, actionable insights. As we brainstormed, we focused on the key challenges potential business owners might face, particularly the need to understand their competition and neighboring businesses in detail. This led us to develop our map feature, offering users a visual representation of the area and helping them strategize for success in their ventures.

What it does

Our tool revolutionizes the way one can search for and evaluate options for commercial real estate. Users can input a location they are looking to buy in, and view a list of potential competitors within a certain radius. This will allow them to compare their projected success based on current data from similar businesses. A frequently overlooked predictor of business success is the number of fraudulent insurance claims by businesses in the surrounding area. Our software also includes an analysis of recent insurance claims to provide an indication of the number of fraudulent claims, providing another metric that your clients can consider.

For the convenience of the user, we have implemented a chat feature that takes in user provided information like location, radius to consider, and business type, and generates local competition. Using a Large Language Model, we have analyzed competitor data to output a succinct and easy-to-read report that contains the information that should be considered whether or not to purchase a property.

How we built it

Our tool is multi-faceted, hence our engineering approach was as well. We used Figma to create our wireframes. Our first step was to use React.js to develop the UI of our project. We integrated the Google Maps API to allow us to display a map on the interface, and to implement functionality to locate the user as well as competitors based on user-input parameters. We used Flask for our backend, and integrated the Assistance API to enable chat functionality and to generate the report based on the information provided. Additionally, we utilized Microsoft's Project Sophia to train a Machine Learning model to detect fraudulent reports as well as analyze the most important factors to consider when purchasing commercial real estate.

Challenges we ran into

Most of our team was relatively inexperienced with a lot of the technologies we decided to venture into. Figuring out the nuances of the API integration, for instance, rendering objects provided by the API proved to be a challenge for us, especially when we couldn't ensure if an error was an API error or a bug in our own code. We encountered a status code of 500 multiple times, and since this usually indicates a problem with the API itself, we were unsure about how to proceed. We solved this by looking through the Google Maps API and familiarizing ourselves with the types and limits of requests we could make. Also, since we developed our chatbot separately from our map interface, integrating the two interfaces as well as our Flask backend, proved to be challenging due to the overlap between the two frontend features. We were able to resolve this by closely observing the aspects of the front-end and back-end that would interact with each other, and smoothing out any problems.

Accomplishments that we're proud of

A lot of the technologies we used were new to most of us, and we were able to work together to overcome challenges in order to get our project working. Apart from learning about frequently used things like API integration, we also utilized trending technologies like LLMs. The most impressive part of this hackathon for us was the vast amount we learned in this little time, which is something we are very proud of.

What we learned

We learned about API integration with various different APIs and what security issues to consider, as well as integrating LLMs and other tools such as Project Sophia. We also used React and Flask for our frontend and backend, which were new to some of us, so we were able to expand the tech stack we are experienced in.

What's next for VentureVisionary

We want to implement functionality where the user can upload an image of the interior of their property, and our tool can send back an image of what the space can hypothetically look like based on the use case given by the user.