Interactive Voice Widget Web Development
Budget: $30 – $50 USD
We are looking for an experienced JavaScript developer to build an interactive voice widget for our website. This widget will function similarly to voice assistants like those from ElevenLabs, with a focus on seamless interaction between audio playback and user speech.
The core functionality involves a two-step process:
• Playback: The widget will first play a pre-generated greeting audio message to the user.
• Recording: Immediately after or during the playback, it must listen for the user's voice response.
Key Requirements:
• The successful candidate must develop a solution that handles the following:
• Audio Playback: Smooth playback of an initial greeting message from an audio URL.
• Overlap/Interruption Handling: The widget must be able to stop the greeting playback instantly when the user starts speaking (voice activity is detected). This is a critical feature.
• Smart Voice Recording: Initiate recording as soon as the user begins their query.
• Dynamic Recording End: The recording should not have a fixed duration. Instead, it must automatically stop when the system detects 3 seconds of continuous silence after the user's speech. This silence threshold should be easily configurable.
• API Integration: Once the recording is complete, the audio file (preferably in a format like WAV or MP3) must be sent to our backend API server via a POST request.
The core functionality involves a two-step process:
• Playback: The widget will first play a pre-generated greeting audio message to the user.
• Recording: Immediately after or during the playback, it must listen for the user's voice response.
Key Requirements:
• The successful candidate must develop a solution that handles the following:
• Audio Playback: Smooth playback of an initial greeting message from an audio URL.
• Overlap/Interruption Handling: The widget must be able to stop the greeting playback instantly when the user starts speaking (voice activity is detected). This is a critical feature.
• Smart Voice Recording: Initiate recording as soon as the user begins their query.
• Dynamic Recording End: The recording should not have a fixed duration. Instead, it must automatically stop when the system detects 3 seconds of continuous silence after the user's speech. This silence threshold should be easily configurable.
• API Integration: Once the recording is complete, the audio file (preferably in a format like WAV or MP3) must be sent to our backend API server via a POST request.