The Retell Community is where innovators, call center leaders, and AI enthusiasts come together to share ideas, learn best practices, and explore the future of voice automation. Join to connect with peers, get early access to product updates, and collaborate on building human-like AI voice experiences that transform customer service.

Retell AI

I'm using a single-prompt agent and GPT 5.1. For a while now I've been getting a pretty annoying error, where later in a conversation the model will begin asking multiple questions at a time and jumping between them. I've been able to make this a bit better via the prompt, but it appears in 30% of my calls. I'm not sure if it's a model thing, a transcription thing, or a prompting thing. Anyone encountered / fixed this before?

Its a prompting issue. you need to adjust it in a structured way or if the logic is complex switch to multi prompt agent or conversation flow as single prompt  can experience behavioral drift and context confusion as complexity increases.

When you say adjust it in a structured way, what do you mean?

i mean write tasks as steps

## Task
1. Ask for user's name.
wait for user response
2. Ask if user needs a refund...
wait for user response

and add explicit instructions to ask one question at a time