Description
Summary I have a voice agent setup built on Vapi that uses a custom RAG knowledge base via Pinecone for answering caller questions. n8n handles the post call automation and pushes data downstream. The setup is mostly working but has a few issues I need fixed: Vapi is not consistently hitting the RAG endpoint during calls, sometimes falls back to generic responses Pinecone retrieval is returning irrelevant chunks, responses are off topic or incomplete n8n workflow is dropping call data intermittently, not all calls are being logged correctly I already have the infrastructure in place. Not looking for someone to build from scratch, just need someone who understands this stack and can diagnose and fix these specific issues. Please only apply if you have hands on experience with Vapi, Pinecone RAG pipelines, and n8n. Looking to get this resolved quickly. Budget is fixed at $175.