Stream Agent
Everything said on stage, working for the event.
The Stream Agent analyses the audio or video stream of your event — in real time, or after it ends — and feeds what was said to Engage, Answer and Protect. Better answers, sharper polls, and questions answered with what was actually presented.
“The hybrid policy applies globally from next quarter. Regional leads will confirm the local detail.”
Q Does the hybrid policy apply globally?
A Yes — from next quarter, with local detail to follow.
“Will this apply to my region too?”
Yes — the policy applies globally from next quarter. Your regional lead will confirm the local detail.
Source · said on stage at 10:02Watch the Stream Agent at work.
A town hall in under two minutes. The Stream Agent listens to the stage and turns what is said into a Q&A pair, an answer for questions already waiting, and an update worth sending. It never publishes: it hands each one to Answer or Engage, and the team approves.
The part of the event nobody types.
The audience’s questions, votes and reactions are already captured. What the speakers say — the answers, the announcements, the detail behind them — usually is not. The Stream Agent listens to it, so the rest of ReactLive can use it.
- Audio
- The speaker’s microphone or the programme feed.
- Video
- The stream or the recording of the session.
- In real time
- While the event is live, so the agents can act on what was just said.
- Afterwards
- From the recording, so answers and the record are complete even when nothing was analysed live.
From what was said to something the event can use.
Q&A pairs
Every question a speaker answers on stage becomes a clean question and answer pair — ready to reuse for repeats, recaps and future events.
Answers to questions already in the queue
When a speaker covers something the audience has asked, the answer is drawn from what was presented and attached to those questions.
New polls and updates
A topic on stage becomes a poll worth running, or an update worth sending — suggested at the moment it is relevant.
Context for every agent
A running record of what has actually been said, so the agents work from the event as it happened rather than only from material prepared beforehand.
It makes Engage, Answer and Protect better.
The Stream Agent does not act on the audience itself. It feeds the other agents — and they still work inside the Off, Assist or Auto mode you set, with your team approving.
Engage
Suggests new polls and updates from what was just said on stage, instead of only from what the audience typed.
About Engage
Answer
Answers better, and answers more: existing questions get the information presented on stage, and the Q&A pairs become event knowledge.
About Answer
Protect
Reviews contributions with the context of what has actually been said and announced on stage.
About Protect
The answer from the stage, next to the question.
When a speaker covers a question that is waiting in the queue, the moderator sees the answer drawn from what was said, attached to that question.
Nothing reaches the audience by itself. The moderator can publish it, keep it as a Q&A pair, or simply mark the question answered.
“Will the hybrid policy apply to every region, or only head office?”
“It applies globally from the first of next quarter. Regional leads will confirm local details with their teams.”
Illustrative view.
The agent that listens, so the others can act.
The Stream Agent does the listening and the analysis. Engage, Answer and Protect decide what to do with it — so what was said on stage improves the event without anything being published that your team did not approve.
- Analyses audio or video, live or after the event
- Feeds Engage, Answer and Protect — never publishes on its own
- Answers are attributed to what was said on stage
- Turned on per event