Does your use case need a vision agent approach or a traditional workflow?
Ask a follow-up question to continue the conversation