House Rental Search API

Table of Contents

• About
• Getting Started
• Usage
• Contributing

About

This project is designed to provide an API for a house rental service that enables users to find properties based on descriptions. It utilizes a state-of-the-art language model from Vertex AI to extract key information such as the number of rooms, location, and rent from natural language descriptions. The API then queries a pre-populated SQLite database to find matching properties based on the extracted data.

Getting Started

These instructions will get you a copy of the project up and running on your local machine for development and testing purposes.

Prerequisites

Before you begin, ensure you have the following installed:

• Python 3.8 or higher
• FastAPI
• Uvicorn, an ASGI server for running FastAPI
• SQLite3 for the database
• vertexai for integrating with Google Vertex AI

You can install the necessary libraries with pip:

pip install fastapi uvicorn sqlite3 vertexai

Additionally, you will need to obtain a credentials.json file for accessing Google Cloud services:

• Download the credentials.json from your Google Cloud Platform console.
• Place the credentials.json in your project directory, or a secure location of your choice.
• Download credentials.json here

Installing

Follow these steps to get your development environment running:

1. Clone the repository to your local machine:

git clone https://yourrepository.com/yourproject.git
cd yourproject

2. Install the required Python packages:

pip install -r requirements.txt

3. Set up the environment variable for Google Cloud credentials:

export GOOGLE_APPLICATION_CREDENTIALS='path/to/your/credentials.json'

4. Initialize the database with the seed data:

python seed.py

5. Start the FastAPI server:

uvicorn main:app --reload

This will start the server on http://127.0.0.1:8000. You can interact with the API using this base URL.

Usage

To use the API, send HTTP requests to the appropriate endpoints:

• Zero-Shot Prompting Endpoint:

  • POST /search_properties/zero_shot/
  • Description: Pass a property description and receive property matches without prior examples.

• Single-Shot Prompting Endpoint:

  • POST /search_properties/single_shot/
  • Description: Pass a property description with a single example to guide the model.

• Few-Shot Prompting Endpoint:

  • POST /search_properties/few_shot/
  • Description: Pass a property description along with multiple examples to help the model understand the context better.

For example, to query the few-shot endpoint, you might use a tool like curl:

curl -X 'POST' \
  'http://localhost:8000/search_properties/few_shot/' \
  -H 'accept: application/json' \
  -H 'Content-Type: application/json' \
  -d '{
  "description": "Searching for a 3-bedroom house in Lagos with a budget of N3,500,000."
}'

This sends a description to the API and retrieves matching properties based on the specified criteria.