Prerequisites Before chatting, you need an agent with knowledge configured. Follow the Create your Hypergraph guide first to get your agent_id.
Chat via the SDK API Send a message and get a response:
import os import uuid import requests API_KEY = os.environ[ "RIPPLETIDE_API_KEY" ] BASE_URL = "https://agent.rippletide.com/api/sdk" headers = { "x-api-key" : API_KEY , "Content-Type" : "application/json" } agent_id = "your-agent-id" # from the agent creation step conversation_id = str (uuid.uuid4()) # one ID per conversation session response = requests.post( f " { BASE_URL } /chat/ { agent_id } " , headers = headers, json = { "user_message" : "What products can I order?" , "conversation_uuid" : conversation_id }) print (response.json()[ "answer" ])
Use the same conversation_uuid for follow-up messages in the same conversation. Generate a new UUID to start a fresh session.
API Reference : Chat API LangChain Integration This feature is experimental.
You can use your Rippletide agent directly as a LLM in LangChain and LangGraph. This lets you replace your current LLM (e.g. ChatGPT) with a hallucination-free Rippletide agent in a few lines.
Installation pip install langchain langchain-openai
Usage Rippletide exposes an Azure OpenAI-compatible endpoint, so you can use AzureChatOpenAI from LangChain:
import os import uuid from langchain_core.messages import SystemMessage, HumanMessage from langchain_openai import AzureChatOpenAI API_KEY = os.environ[ "RIPPLETIDE_API_KEY" ] agent_id = "your-agent-id" # from the agent creation step conversation_id = str (uuid.uuid4()) rippletide_llm = AzureChatOpenAI( model = "v1" , api_key = API_KEY , azure_endpoint = "https://agent.rippletide.com" , azure_deployment = "v1" , api_version = "2024-12-01-preview" , openai_api_type = "azure" , default_headers = { "x-rippletide-agent-id" : agent_id, "x-rippletide-conversation-id" : str (conversation_id), }, ) messages = [ SystemMessage( content = "You are a helpful assistant." ), HumanMessage( content = "What products can I order?" ) ] response = rippletide_llm.invoke(messages) print (response.content)
Use in a LangChain chain Once initialized, rippletide_llm works like any other LangChain LLM:
from langchain.prompts import ChatPromptTemplate prompt = ChatPromptTemplate.from_messages([ ( "system" , "You are a helpful assistant for an electronics store." ), ( "human" , " {question} " ) ]) chain = prompt | rippletide_llm result = chain.invoke({ "question" : "How long does delivery take?" }) print (result.content)
Next Steps
Evaluate your agent Test for hallucinations before deploying to production
Full Developer Guide Complete API reference with advanced examples
