iFetch Chrome Extension

The iFetch Chrome Extension is a project that integrates the iFetch conversational agent into the Farfetch store, allowing users to interact with a chatbot while browsing the store's items.

Installation

To install the extension, follow these steps:

1. Clone the project repository.
2. Open Google Chrome and navigate to the Extensions tab.
3. Select the "Manage Extensions" option, which will open a new tab displaying all installed extensions.
4. Click on the "Load Extension" button, which will open a new window.
5. Select the "Build" folder from the iFetch Chrome Extension project.
6. The extension will be installed and ready to use.

Usage

To use the iFetch Chrome Extension, follow these steps:

1. Open the page of an item in the Farfetch store.
2. Click on the icon of the iFetch Chrome Extension, which will open a chat window.
3. Interact with the chatbot to ask questions or get more information about the item.

Editing

If you wish to extend or customize the current version of the project or use it as an interface for your own project, please follow these instructions.

Change Endpoint

The first step is to change the endpoint used by the project. To do this, you need to modify the MESSAGES_ENDPOINT field in both the "App.js" and "Content.js" files. Update the endpoint to the URL of the desired server.

Send Messages

The messages are sent to the backend with the following JSON format:

{
  utterance : string,
  user_id : string, 
  session_id : string,
  user_action : string,
  interface_selected_product_id: string,
  image: base64 byte dump
}

The utterance field contains the message to be sent, the user_id field contains the unique identifier for the user, the session_id field contains the unique identifier for the session, the user_action field contains the action taken by the user, and the interface_selected_product_id field contains the identifier of the selected product. After making this changes you have to build the project using npm or yarn, and running the command yarn run build.

The interface expects responses in the following JSON format:

{
  has_response : bool,
  recommendations : List<Recommendations>,
  response : string,
  system_action : string
}

Where recommendations are presented in the following JSON format:

{
  brand : string,
  description : string,
  id : int,
  image_path : string
}

Thank you for using the iFetch Chrome Extension! If you have any questions or feedback, please feel free to contact us.