Set up your conversation copilot
Choose the languages, attach the documents you might need to cite, and pick the audio Contexa should listen to.
Checking the Contexa API…
Languages
AssemblyAI Universal-3.5 Pro Realtime speech_model=universal-3-5-pro
Low-latency streaming speech-to-text with native code-switching across 18 languages, including English and Japanese. Speech is transcribed as spoken, then translated into Indonesian.
Suggested answers
Clear and polite, for judges or clients. You can redraft any answer in another style from the copilot.
Audio source
Streaming diarization tags each turn with who said it.
No context documents yet
Answers can only cite what you attach. A README, proposal, or architecture doc works well. Add them in the Context panel.
Your browser will ask which tab to share. Pick the one with the webinar and turn on “Share tab audio”.