Skip to content
Nappai Documentation

ConversationChain

This component lets you create chatbots or virtual assistants within your Nappai automation workflows. It uses a language model (like those from OpenAI or Google) to understand and respond to user input. Importantly, it can also remember previous parts of the conversation, making the interaction feel more natural and coherent.

Relationship with Langchain

This component uses the Langchain library, a powerful tool for building applications with large language models. Langchain handles the complex details of managing the conversation and the language model, allowing you to focus on building your application.

Inputs

  • Input: This is where you provide the text input for the AI to process. Think of it as what the user types into the chatbot. This field is required.
  • Language Model: Select the language model you want to use for your conversation. Nappai supports various models; choose the one best suited for your needs. This field is required.
  • Memory (Optional): This allows the AI to “remember” previous parts of the conversation. If you leave this blank, the AI will treat each input as a completely new conversation. If you select a memory option, the AI will maintain context across multiple turns in the conversation.

Outputs

The component produces a text message as output. This message is the AI’s response to the input you provided. You can use this output to display the AI’s response to the user or feed it into other components in your Nappai workflow.

Usage Example

Imagine you’re building a customer service chatbot. You could use the ConversationChain component like this:

  1. Input: A customer types “I’m having trouble logging in.”
  2. Language Model: You’ve selected a powerful language model trained for customer support.
  3. Memory: You’ve chosen a memory option to store the conversation history.
  4. Output: The AI responds with “I understand. Could you please provide your username?” Because of the memory, if the customer then provides their username, the AI can use that information in subsequent responses.

Templates

[This section would list specific Nappai templates where this component is used. Since no templates were provided, this section remains blank.]

Several components work well with ConversationChain:

  • Chat Input: This component can be used to get user input for the ConversationChain.
  • Chat Output: This component displays the output from the ConversationChain to the user.
  • Conversation Chat Memory: This component provides memory functionality for the ConversationChain, allowing it to remember previous interactions. (And similarly for other memory components listed in the provided data).
  • Various Language Model Components: These components provide the language models that power the ConversationChain.

Tips and Best Practices

  • Choose the right Language Model: Select a language model appropriate for your application’s needs. Some models are better suited for specific tasks, such as customer service or creative writing.
  • Use Memory for Context: Using memory significantly improves the quality and naturalness of the conversation.
  • Test Thoroughly: Test your ConversationChain with various inputs to ensure it behaves as expected.

Security Considerations

  • Data Privacy: Be mindful of the data you feed into the language model. Avoid including sensitive information unless absolutely necessary and the model is appropriately secured.
  • Model Selection: Choose reputable language models from trusted providers to minimize the risk of malicious or biased outputs. Review the security and privacy policies of the chosen model provider.