Coach Audio - Analyze Your Voice Like a Pro With Our Advanced AI-Powered Tool!
Record your pitch, upload an audio file or paste the transcript. Get structured feedback on WPM, filler words, clarity, value proposition and much more. Privacy-first: audio in the browser, self-hosted Ollama AI.
Privacy:. The audio remains in your browser and is never sent to external servers. Transcribed text is sent to the Ollama self-hosted server on Hetzner VPS for AI analysis. No data is saved.
Register your pitch
Speak clearly for at least 30 seconds. Transcription occurs live within the browser.
Come utilizzare Pitch Coach Audio
Choose Input Mode
Record a pitch from the microphone (up to 60 seconds), upload an audio file, or paste the transcribed text directly.
Speak or transcribe the pitch
In registration mode, transcription happens in real-time with the browser's Web Speech API (requires Chrome or Edge); otherwise paste pre-written text.
Wait for analysis
Metrics are calculated immediately in the browser; AI feedback comes from the Ollama self-hosted server.
Read score and feedback
Review overall score, evaluated dimensions (clarity, value proposition, problem-solution fit, ask, energy), strengths and areas of improvement.
Suggerimenti
- Speak for at least 30 seconds to get more accurate WPM and filler word statistics.
- In the "Stick Text" mode, the duration is estimated at an average rate of 130 words per minute: for a real pitch record with the microphone.
- Review the "Areas for Improvement" section before repeating the pitch: it's generated by the AI Ollama model based on both metrics and real content of the transcription.
Domande frequenti
Does audio recorded by me go to a server?
No, the audio stays in your browser and is never uploaded elsewhere. Only the transcribed text is sent to the Ollama self-hosted server on VPS Hetzner for AI feedback generation; no data is saved.
What happens if my browser doesn't support speech recognition?
Web Speech API is mainly supported by Chrome and Edge. If it's not available, the tool will signal this and you can still analyze the pitch using the "Stick Text" mode with manual transcription.
What does WPM measure (words per minute)?
Measures the speech rhythm and pitch with a qualitative label (e.g. "Excellent") calculated against predefined thresholds: low WPM speeds slow attention, high speeds make it hard to follow.
What happens if AI feedback is not available?
Metrics remain visible even if the Ollama server doesn't respond: only qualitative aspects (score, size, strengths and areas of improvement) are missing with an explicit message indicating this.
Can I load a pre-recorded audio file instead of using the microphone?
Yes, you can upload audio files (mp3, wav, webm) up to 5 MB. Automatic transcription of files requires support for the Web Speech API on the audio element: if it doesn't work in your browser, copy the transcribed text and use the "Paste Text" mode.