Reca - a real estate conversational agent

This App was created for LabLabAi's Llama 2 Hackathon with Clarifai. This conversational agent take a user's description of their ideal home and finds the best results in our database and returns a detailed description of the property. FAQ How to use App? Just describe the type of property you are looking for... User: "I am looking for a house in Vilamoura with 3 bedrooms and 3 bathrooms" Bot: "I have found 5 properties that match your criteria. Here is the one of the properties: Townhouses in Private Complex. Located in Vilamoura. Has 3 bedrooms. Has 4 bathrooms. With a 161 m2 build area. Here is the link: (link)" User:"Please describe the property to me" Bot: "Luxury tourist resort in Vilamoura, Algarve with fully furnished apartments and townhouses, surrounded by nature reserve and close to the beach. Offers communal pool, Cafe, Kid's club, and concierge services." User: "Please show me some pictures" Bot:"Here are some photos of the property:(shows pictures)" What tech stack is being uses? The ui was built in Python using the Streamlit library. For the conversational part of the app, we are using Meta's Llama 2 through a workflow created on the Clarifai platform. We also created a demo database with information about several properties using MongoDB Atlas, but we intend on integrating a more mainstream API from a well know real estate portal, such as Idealista. How to test the APP? Follow the Github README.md to run the project locally.