Working with Projects

Projects are how you organize and share work in alwaysAI.

Creating a New Project

To create a new Project, log into alwaysAI and go to your Dashboard. In the Projects view select the red plus (+).

You have the option to create a Project that includes one of our starter applications or to create an empty project and add assets later.

To configure the Project you’ve just created on your Development Computer simply create an empty directory:

$ mkdir app && cd app

and run the CLI command:

$ aai app configure

From here you’ll be prompted to select the project you’re going to use and the target you’d like to install it on.

You can follow the same process to configure an older alwaysAI application directory as a working Project. This is useful if, for example, you have already downloaded the alwaysAI starter applications from the CLI. Just run the aai app configure command from within any chosen directory, pick a name, and it will be ready to go.

Running Your Application

Install the Application

You can run your application locally on your computer, or remotely on an edge device.

For detailed instructions see our guide on starter applications.

To get started quickly on your LOCAL COMPUTER, run the following command:

Note: This assumes you have selected “Your local computer” when running the aai app configure command above.

  • On mac or windows

    $ aai app install
    
  • On linux

    $ aai app deploy
    

This command prompts you for your alwaysAI account credentials and then installs the application’s models and a Python virtual environment (venv) into the application directory.

Start the Application

Now start your application with the command:

$ aai app start

The log output from app.py is forwarded through the CLI to your terminal, including the labels that the model has been trained to recognize.

Once the app starts, a log will be printed with the link:

[INFO] Streamer started at http://localhost:5000

Open this URL in a browser to see the video and text data generated by the app. To stop the application, simply hit CTRL-C or click the red “stop” button on the web page.

Add Models to a Project on the Web

From the web, you can add a model to your project from the alwaysAI Model Catalog or using your own private models in your Dashboard.

Public Models

On the web, simply go to the Model Catalog. and select a model. Within the model details will be a card with instructions on using the model; select your project from the drop down at the bottom of this card. That’s it!

Private Models

Any private models you’ve uploaded to your account (or your collaborating on in another users project) will be listed in your Dashboard under My Models.

You can add a private model to a Project by clicking through to the model’s detail page and selecting your project from the drop down (similar to adding a public model) or by clicking the red plus (+) to the right of the model section on your project’s detail page, then selecting the model you’d like to add to your project from the drop down.

See here for additional detail on adding a model to your project.

Add Collaborators to a Project

Only the people specifically added to a project, Collaborators, can use it. This ensures only you determine who can use your private models and devices that run the application.

To add Collaborators to a Project you own, find the Collaborators card on the project detail page and select the red “+” button. You’ll need to know the email address of the Collaborator you invite; if they do not have an alwaysAI account already we’ll send them an email to sign up and add them to your project automatically.

Add Devices to a Project

You can add a device to your project by opening a command prompt, changing to your project directory:

$ cd app

and then executing the command:

$ aai app configure