Support

AI Connector

GenAI world has significantly changed over the last year. If you are a Software Engineer or CS Researcher, you likely have experience building AI agents using tools like Langchain, Llamaindex, Cohere, etc.  However, building an AI agent logic is only half of the work. Now you have to deploy it, connect communication channels like Web-widget or WhatsApp, be able to track recent chat histories, see Stats, send email notifications, and many more. 

That's why we built AI Connector

Configure your Endpoint as following

# Input parameters (json):

messages - list of the latest messages. Each message has a role: user/assistant, content

user_id - id of the user

channel - channel id.

Example: {"messages":[{"role": "user", "content": "hello"}], "user_id":1, channel:5}


# Response parameters (json):

messages - list of generated responses that includes:

      content - text response. Ex, "Hello, how may I help you?"

      buttons (optional) - list of buttons. Ex, ["talk to agent", "get a quote"]

      sources (optional) - list of sources with link. Ex, [{"link":"https://ailabs.us", "title":"source website"}]

      action (optional) - name of action to call. For example, "connect-human-agent" will connect with human agent.

      inputs (optional) - input parameters to fill out the user infomation in the Inbox. Ex, [{"name":"first_name", "value":"Chris"}].  

Output response example: { "messages":[{"content": "hello", "buttons": ["connect agent"]}] }


We are eager to hear from you. Please reach us out at success@webapi.ai for a feedback or suggestions.

Did this answer your question?
We're glad you could find the answer you were looking for
Your feedback has been recorded. If you'd like to tell us how we can improve this content please fill out the form below.