Speech-to-Text for 1-to-1s
Overview
Let your voice do the writing. Learn Amp now supports speech-to-text input for free-text questions in 1-to-1s, helping users respond faster and more naturally—powered by AssemblyAI.
This feature allows users to speak their responses to free-text 1-to-1 questions. Their spoken words are automatically:
Transcribed in real time
Polished for clarity (e.g. filler words like "um" and "uh" are removed)
Ideal for users who prefer verbal communication or want to save time when completing longer-form answers.
Prerequisites
To enable speech-to-text in 1-to-1s, Admins must:
Toggle ON the setting under: Company Settings → AI Features → AssemblyAI Speech-to-Text for 1-to-1s
Note: The company setting is OFF by default. Admins must turn it ON for the feature to take effect.
Quick-Start Guide
Step 1: Enable the Company Setting
Go to:
Company Settings → AI Features → AssemblyAI Speech-to-Text for 1-to-1s
Toggle the setting ON.
Step 2: Start a 1-to-1 Form
Navigate to a 1-to-1 that includes:
Free-text questions, or
Opinion scale questions with free-text follow-ups
Step 3: Use Speech-to-Text
When filling in the form:
Tap the microphone button next to a free-text field
Speak your response
Click "Stop" once finished
Your voice will be transcribed and cleaned up automatically
You can review and edit the text before submitting.
Remember to save your draft or submit your 1-to-1 once you are satisfied with the responses you've added.
Data Retention and Model Training
Where is our data processed?
All transcription processing is handled through AssemblyAI's services located in Ireland, EU. This ensures compliance with UK and European data protection regulations and keeps your data within the EU jurisdiction.
How long is our data stored by the transcription service?
Learn Amp has implemented Zero Data Retention (ZDR) with AssemblyAI. This means that voice recordings and generated transcripts are instantly deleted from AssemblyAI's servers once the processing is completed, ensuring your data privacy and security.
Is our data used to train AI models?
No. We have a contractual agreement with AssemblyAI that explicitly prevents our data from being used to train or improve their models. Your data remains confidential and is not used for any machine learning purposes.
FAQs
Q: Which type of questions support this?
A: The speech-to-text button appears for all free-text questions in 1-to-1 forms and any free-text follow-ups to opinion scale questions.
Troubleshooting
Issue | Solution |
|---|---|
Speech-to-text button not showing | Ensure the company setting are enabled |
Transcription isn't working / nothing happens when I start recording | Check microphone permissions in your browser/device |
Text doesn't appear polished | Try re-recording your answer: AI filters filler words, but may struggle with background noise |