An assistant is the main functionality of the chatbot, it helps the user in a certain task programmed from an instruction. 

We can visualize our assistants in a table:

There, we can filter by the model, and synchronize our assistants if we have changes in the Open AI web that we want to get.

Clicking the  “+ assistant” button opens a modal with the settings of the assistant creation.

• Add a label to identify the assistant.
• Add a description to help the user understand what the assistant does.
• Add a programmed first message to the assistant when the user starts the conversation.

Select the AI model of the assistant to perform their task, take note that while the gpt-4 models are better for performing, answers may take longer loadings between requests and in some cases can cause problems if the answer is too complex.

Add instructions to the assistant to perform better in their response. For example, if the assistant needs to assist in business sales, then the instruction needs to be sometime like: 

“Take the role of a seller assistant, your duty is helps the customer in questions related about our products”

Remember, the better the instructions are, the better the response will be.

Select the storage, previously created in the storages tab, that the assistant will use to improve their response.

Tweak the visibility for the assistant, select the condition to show the assistant in different devices.

Select if the assistant needs to be included or excluded for target (Home, Blog, Search, 404), posts, pages, categories and tags

Vector store or Storage refers to the method of storing data in a high-dimensional vector space. When data (such as text, images, or other forms of information) is processed by an AI model, it is often converted into vectors, which are arrays of numbers that represent the features of the data in a way that the model can understand and work with. These vectors are processed to generate meaningful responses.

This table displays our storages uploaded in Open AI and two buttons to synchronize or create a storage.

Synchronize: get storages that were uploaded in Open Ai’s website and add it to the table.
+ Storage: create a new storage. Note that creating a storage will upload to Open AI automatically.

As for the storage, we can see:
Label.
Description.
Assistants: the assistants that use the storage.
Files: files saved in the storage.
Status: if the file was deleted from the Open AI it’s marked as deleted, you have the option to delete it or upload again.

Lastly the are three actions to interact with the storage:

• Upload: when the storage is deleted from Open Ai, we can upload again clicking in Upload.
• Edit: edit the storage.
• Delete: delete the storage.

• Add a label to identify the storage.
• Add a description to help the user understand what the storage does.

Select a file, previously created in files tab, to add it into the storage.

Adding a file to a vector store allows for efficient information retrieval. The file is processed, chunked, embedded, and stored in a vector database, enabling accurate keyword and semantic searches for better, contextually relevant results.

This table displays our files uploaded in Open AI and three buttons to synchronize or upload files.

• Synchronize: get files that were uploaded in Open Ai’s website and add it to the table.
• + Post Type File: create a file with content selected from the post type (All, pages, posts)
• + Extra Files: upload a file from the computer.

Note that both actions, + Post Type File and + Extra Files, upload the file to the Open AI automatically.

As for the file, we can see:

• Label.
• Description.
• Post types.
• Origin: external means that it was created in Open Ai’s website, and internal, it was created from the copilot plugin.
• Assistants: the assistant that uses the file.
• Approximate cost: for more info about costs here.
• Status: if the file was deleted from the Open AI it’s marked as deleted, you have the option to delete it or upload again.
• Date

Lastly the are three actions to interact with the file:

• Upload: when the file is deleted from Open Ai, we can upload again clicking in Upload.
• Download: download the file.
• Delete: delete the file

• Add a label to identify the file.
• Add a description to help the user understand what the storage does.
• Select the post type for file generation.

• Add a label to identify the file.
• Add a description to help the user understand what the storage does.
• Select the post type for file generation.