I’ve worked with Retell + Pinecone on multiple voice AI projects, and slow responses usually come from the vector query or streaming setup—not Retell itself.
I can help you pinpoint the exact bottleneck and optimize latency so replies feel real-time again.
Before we jump in, are you using Pinecone serverless or standard, and how big is your vector dataset?
@deankaks