An interactive way to search flats in Vienna using Microsoft AI services.
Clicking on button bellow will redirect you to the Azure portal to deploy the resources necessary to conduct this demo using the ARM template provided in this repository.
This script will create the following resources: Azure AI Search, Azure Cosmos DB, Storage Account, Azure Maps Account, Azure OpenAI Service (with the deployments of gpt-4o-realtime-preview
and text-embedding-3-large
)
-
Clone the repository.
git clone https://github.com/moalsayed95/domago.git
-
GitHub Codespaces is a cloud-based dev environment that lets you code anywhere, with pre- configured setups directly from GitHub repositories. Access it via browser, VS Code, or GitHub CLI.
Please select your forked repository from the dropdown and choose
default_environment (Python 3.11)
as your default codespace configuration -
Log in into your Azure account using the Azure CLI. Use
az login
to start. -
After deploying the resources, you will need to configure the environment variables in the .env file. This file is automatically created running the following command within the terminal in your Codespace:
./get-keys.sh --resource-group <resource-group-name>
If you run into into permission issues, run:
chmod u+r+x ./get-keys.sh --resource-group <resource-group-name>
-
Create the AI search Index and index the dummy data from the
data
folder. The data folder has the flats data in JSON formatdata/flat_cata.json
. You can do this by running theindex_manager.py
script in the backend project. -
Run the backend server.
cd app/backend && python app.py
-
Open another terminal and run the frontend server.
cd app/frontend && npm run dev
Both the frontend and backend servers should be running for the app to work.
I am looking for a flat in the city center show me some options
I like this flat (say the flat's title and ask for more details)
I like this flat (say the flat's title) could you add it to my favorites
I like this flat (say the flat's title) could you zoom in on the map