Skip to main content

Generative AI in ConvyAI - LLM-Powered Bots and AI Agent Orchestration

Overview​

ConvyAI integrates Generative AI directly into the conversational platform through two complementary approaches, both built on the same foundation of AI Models and LLM providers.

Knowledge Base - A RAG-based approach where bots retrieve relevant content from an indexed knowledge base and pass it to the LLM to generate grounded, brand-aligned responses. Best suited for conversational use cases where accuracy and source grounding are critical.

AI Agent - A process-oriented approach where the AI Agent node orchestrates interactions between the LLM, images, MCP servers, and custom JavaScript functions. Best suited for automation scenarios that go beyond conversation: data extraction, classification, multi-tool workflows, and complex decision-making.

Both approaches share the same AI Models configuration, which defines the LLM provider, model version, and inference parameters.

Architecture​

In this section​

Shared infrastructure​

  • AI Models - Configure the LLM providers used by both approaches.

Knowledge Base​

  • Knowledge Base - Understand the RAG indexing pipeline and knowledge base structure.
  • Settings - Configure API credentials and Redis connection for the indexing service.
  • Indexing - Create and run indexing batches from S3 or website sources.
  • Document Management - View, search, and remove indexed content.
  • Generative AI Bot - Use the Knowledge Base node in the FlowDesigner.

AI Agent​

  • AI Agent - Understand the AI Agent concept and when to use it.
  • AI Agent Node - Configure the node fields, prompt, tools, and interaction settings.
  • Image Analysis - Attach and analyze images within the agent prompt.
  • MCP Servers - Connect external tools and services via Model Context Protocol.
  • Callable Functions - Define custom JavaScript functions the AI can invoke during execution.