Real-time audience feedback
See exactly which sentence lands and which one flops.
Speakers stream live audio. Listeners tap green or red. We transcribe the talk into sentences and pin every reaction to the moment it was said. You walk off stage knowing what worked.
How it works
Speaker streams
From any browser. We capture audio with the built-in mic. No extra hardware. Speakers see the audience pulse live.
Listeners react
Audience opens a code or QR. One tap green for agreement, red for confusion or disagreement. No account required.
You get insights
Each sentence is transcribed and annotated with its reactions. Best moments and weak moments surface automatically.
Built for
Conferences & talks
Know which slide actually moved the room and which one lost everyone.
Lectures & training
Spot the explanation that confused half the class so you can revisit it next session.
Music gigs
Hear which song the crowd loved beat by beat. For set-list tuning between cities.
Webcasts & remote panels
Distributed audience, real-time signal. No conference-call awkward silence to interpret.
Security & data handling
Live audio plus identifiable reactions is sensitive. We are upfront about what we do with it.
What we record
Audio while an event is live, plus the transcript and the reaction stream. Audio is deleted immediately afterwards.
Where it is stored
EU region (Germany) by default.
Retention
Raw audio is immediately auto-deleted. Transcripts and reactions stay until the speaker deletes the event.
Your rights
Listeners can delete their reactions and notes at any time. Speakers can also export and delete all event data whenever they want.
Full details: Privacy Policy · Terms of Use
FAQ
- Do listeners need to install an app?
- No. They open a link or scan a QR code, tap green or red. Optional display name, no signup required.
- Is the speaker's audio shared with the audience?
- No. Listeners hear the speaker in person (or via a separate live stream they already have). We are focused on the feedback loop, not on being a streaming platform.
- How accurate is the transcription?
- We use cloud speech-to-text providers (Deepgram by default) tuned for talks. Quality is best with a decent mic and minimal background noise.
- What does an event cost?
- Free for listeners; pricing details for Speakers before public launch. Speakers being signed up now are free forever.
- Can I delete an event afterwards?
- Yes. From the speaker dashboard.
Something else? yeysus@gmail.com
About
FeedBackApp started as an idea sometime around 2011, out of a personal frustration: Jesus was notoriously bad at writing the so-called meeting minutes. He was leading some large projects, and he was expected to send out the notes with summarized conversations and action points. Back in the day, people actually read those minutes, and sometimes it came back: “I actually meant it like this…”, or worse, “I never said that”. After every meeting he would spend hours re-writing and reconstructing a “politically correct” version of who said what, which decisions had been made, and what everyone had agreed to do. It was a tedious manual process. It screamed for automation.
The original idea was very simple: the app should capture conversations sentence by sentence, and at the end the participants would agree on the “decisions” and “action points”. Then everyone would receive their meeting minutes by email, with clear action points. Over the years we made several attempts to build it, among others with IBM Watson, but the technology was never really satisfying. Speech recognition was too inaccurate, and getting the spoken words sentence by sentence was a nightmare. Add to that a corporate environment full of international people and many different accents, and it never made it into a real product.
November 2022 changed everything. OpenAI and ChatGPT took speech-to-text to the next level and made it available to everyone. The new large language models (LLMs), together with GenAI-assisted software development, so-called “Vibe Coding”, a term popularized in February 2025 by OpenAI co-founder and computer scientist Andrej Karpathy, have dramatically lowered the time and effort it takes to build software. A small team can now achieve in weeks what large teams with many specialized roles used to need years for.
FeedBackApp is the result of that journey. It started as a way to solve a personal meeting-minutes problem and has grown into something bigger: a platform that captures spoken words and helps people see which ideas resonate and which ones do not.
We are a small operation based in Berlin.
If you would like to talk to a human about the product, write to yeysus@gmail.com.
We love history and Wikipedia. You can read more about these topics on the English Wikipedia:
- Vibe Coding: https://en.wikipedia.org/wiki/Vibe_coding
- Large Language Model: https://en.wikipedia.org/wiki/Large_language_model
- OpenAI: https://en.wikipedia.org/wiki/OpenAI