I'm developing an agent framework(langgraph/other options) that integrates voice capabilities(vapi), but I'm unsure about the best approach to configure voice interactions between both. So far the only approach I see is tools via REST , but that is limited to only dashboard to setup and not via code. Specifically, I'd like to know if: 1. **Configuration via SDK**: Is it possible to configure voice interactions programmatically using the VAPI SDK? 2. **Integration with Language Models**: How can I integrate language models like LangChain or LangGraph with VAPI to enhance voice interactions?