### Answer The issue you're experiencing with delayed transfer_call execution is related to how LLMs interpret tool calling instructions. Here are the recommended fixes: 1. Add Explicit Tool Usage Instructions The Prompt Engineering Guide emphasizes defining clear triggers for tool usage [(1) ]() . Your current prompt needs more explicit instructions about when to call the function. Update your prompt to include:

## Tool Usage Instructions
When customer requests to speak with a human or customer service:
1. Say EXACTLY: "I completely understand. I'll connect you with our customer service team now."
   - If customer is speaking in a language other than English, translate this message naturally into their language.
2. IMMEDIATELY call the transfer_call function - do NOT wait for user response
3. Do NOT ask follow-up questions before calling transfer_call
4. Do NOT offer alternatives like support tickets before transferring

2. Configure Speech Behavior Settings The transfer_call tool has specific speech behavior configurations that control when the agent speaks [(2) ]() . Ensure you've configured: - **Speak during execution**: Enable this so the agent speaks the transfer message when the function is called - **Speak after execution**: Disable this to prevent the agent from continuing conversation after initiating transfer These settings control the agent's behavior when the user is not talking during the function call period . 3. Use Tool Call Strict Mode If you're using a single/multi-prompt agent, enable

tool_call_strict_mode

in your LLM configuration [(3) ]() . This helps ensure more reliable function execution. 4. Avoid Conditional Language Remove phrases like "Would you like me to..." or "I can help by..." from anywhere near your transfer instructions. The LLM may interpret these as requiring user confirmation before proceeding [(1) ]() . The key issue is that without explicit instructions telling the LLM NOT to wait or ask follow-ups, it may default to conversational behavior instead of immediate tool execution .

​ To learn more, see the sources I considered: [[1] Prompt Engineering Guide](https://docs.retellai.com/build/prompt-engineering-guide) [[2] Transfer call](https://docs.retellai.com/build/single-multi-prompt/transfer-call) [[3] Update Retell LLM](https://docs.retellai.com/api-references/update-retell-llm) ​

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

### Answer You're correct - I apologize for the confusion in my previous response. Transfer Call tools only have "Speak During Execution" settings, not "Speak After Execution" [(1) ]() . For Tool Call Strict Mode in the UI: When creating or editing your agent in the dashboard, you'll find the Tool Call Strict Mode setting in your LLM configuration [(2) ]() . This setting is available when using certain supported models and helps ensure more reliable function execution . Revised recommendations for your transfer issue: 1. Configure "Speak During Execution" - Set this to "Static Text" with your exact transfer message: "I completely understand. I'll connect you with our customer service team now." [(1) ]() 2. Strengthen your prompt instructions - Based on the Prompt Engineering Guide, be extremely explicit about immediate execution [(3) ]() :

When customer requests human assistance:
- Call transfer_call function IMMEDIATELY
- Do NOT ask follow-up questions
- Do NOT offer alternatives
- Do NOT wait for user confirmation

3. Enable Tool Call Strict Mode in your LLM settings if you're using a supported model (like GPT-4.1, Claude, or Gemini) [(2) ]() The 15-second delay you're experiencing suggests the LLM is treating the transfer as optional rather than mandatory. Making the prompt instructions more directive and enabling strict mode should resolve this [(3) ]() [(2) ]() .

​ To learn more, see the sources I considered: [[1] Node Functions](https://docs.retellai.com/build/conversation-flow/conversation-node-functions) [[2] Get Retell LLM](https://docs.retellai.com/api-references/get-retell-llm) [[3] Prompt Engineering Guide](https://docs.retellai.com/build/prompt-engineering-guide) ​

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

Thanks for the follow-up, @soln. I'll get back to you soon.

​ To learn more, see the sources I considered: [[1] Get Retell LLM](https://docs.retellai.com/api-references/get-retell-llm) [[2] Update Retell LLM](https://docs.retellai.com/api-references/update-retell-llm) [[3] Create Retell LLM](https://docs.retellai.com/api-references/create-retell-llm) [[4] List Retell LLMs](https://docs.retellai.com/api-references/list-retell-llms) ​

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