Checklist
Issue or Suggestion Description
I am testing the Echo Ear board using an OpenAI API key. The conversational bot works as expected, including tool calls (e.g., checking time and date).
However, in roughly 90% of the interactions, the board interrupts its own response and begins replying to itself. This seems related to insufficient suppression of its own audio output.
Is there an available workaround or an easy way to improve rejection of the device’s self-generated audio?