Skip to content

Add message to a conversation

Adding messages

The add_user_message() function allows you to add a user message to an existing conversation with a specified agent in Blipper.

This function requires the following parameters:

  • conversation_id (type: str): The ID of the conversation to which the message will be added, provided as a string.
  • message (type: str): The user message to be added to the conversation, provided as a string.
  • agent_id (type: str): The ID of the agent associated with the conversation, provided as a string.

Note

Before using this function, ensure that you have created the conversation by visiting the Create conversation section. Additionally, make sure that the agent has been created. If not, refer to the Create agent section.

1
2
3
4
5
6
7
8
9
from blipper import Blipper

blip_obj = Blipper(blipper_api_key="lqqRYOOnSerIOP973hjskR", model="galileo", verbose=False)

agent_response: str = blip_obj.add_user_message(
    conversation_id="my_conversation_id",
    message="Hi",
    agent_id="my_agent_id"
)

Response

The response returned by the agent is typically in the form of a string or structured data depending on the agent's capabilities. For example, if you're interacting with a Food Agent, the response might include information about food items, menus, or recommendations.

Welcome! How can I assist you today?

Note

The agent may not be able to respond to personalized business questions initially. To enable the agent to respond to such queries, you'll need to add a file containing relevant business information such as prices, business data, services, etc. For instructions on how to do this, please visit the Add file to agent section

Streaming responses

Streaming responses feature is currently not available, but we are constantly working on adding new features and improvements to enhance your experience.

Streaming responses allow for real-time delivery of responses from the agent as they are generated. This feature enables more interactive and dynamic conversations. However, at present, streaming responses are not yet implemented.

Expected Behavior: While awaiting the implementation of streaming responses, the system will wait until the entire response is generated before delivering it. This may result in slightly longer response times compared to streaming-enabled platforms.

Welcome! How can I assist you today?

We appreciate your patience as we continue to enhance our platform to provide you with the best possible experience.