​ To learn more, see the sources I considered: [[1] Subagent Node](https://docs.retellai.com/build/conversation-flow/subagent-node) ​

### Answer The subagent node is used for dialogue with the user while allowing the agent to call tools/functions during the conversation. [(1) ]() It's designed for scenarios where the agent should decide whether and when to use a tool/function based on the conversation context. ## How It Works When a subagent node has tools/functions attached, the LLM receives both the node instruction and the list of available tools/functions. During the conversation, the LLM determines when a tool/function should be called based on context, extracts the required parameters, and invokes it while maintaining the dialogue with the user. Key capabilities include: - Multiple tools/functions can be added to a single subagent node - The agent can continue talking while a tool/function executes - Tool/function results are available to the LLM for generating follow-up responses ## Special Use Cases Subagent nodes are best for context-dependent actions during dialogue. For example, looking up an order only if the user asks for it. When to use subagent nodes vs function nodes: - **Subagent Node**: LLM-driven execution - called when the LLM decides it's appropriate, supports multiple tools per node, intended for full dialogue with tools available - **Function Node**: Deterministic execution on node entry, one tool per node, not intended for dialogue If you want guaranteed execution every time the flow reaches that step, use function nodes. Use subagent nodes when the agent should decide whether and when to call a tool/function based on what the user says.

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer Based on recent updates, here are key improvements to conversation flow capabilities: ## Conversation Flow Components Conversation Flow Components let you build reusable sub-flows and functions that sync automatically across all connected agents. [(1) ]() This feature allows you to: - Build once, use everywhere - Update once, sync instantly across all agents - Keep agents consistent - Save hours of maintenance and rebuild work Components are perfect for security checks, language routing, surveys, and any shared logic. ## Enhanced Subagent Node Features Recent improvements to subagent nodes include: **Node-Level Knowledge Base**: You can now assign a specific knowledge base to individual subagent nodes, allowing agents to retrieve more accurate answers based on context, respond with lower latency by narrowing the search space, and deliver higher customer satisfaction with precise responses. [(2) ]() **Extract Dynamic Variables in Real Time**: Agents can now extract and store dynamic variables live and mid-conversation. Your agents can automatically save names, emails, budgets, and preferences; reference collected data anytime during the call; and pass dynamic variables to webhooks, CRMs, or integrations. **Global Node Fine Tuning**: You can now add example conversations for AI to know when and when not to jump to Global Nodes in your agent's conversation flow. ## Use Case Improvements For complex workflows, the platform now supports Agent Transfer, allowing you to seamlessly switch between Retell agents during a call. This unlocks modular, reusable agents for specific tasks like language routing agents, CSAT survey agents, and payment agents.

​ To learn more, see the sources I considered: [[1] 🚀 Conversation Flow Components are live!](https://community.retellai.com/t/conversation-flow-components-are-live/150) [[2] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.