English Audio Conversation Annotation
Budget: ₹750 – ₹1,250 INR
I am preparing a large batch of English conversation recordings for machine-learning training and need extra hands right away. Your job is to listen to each clip, split or merge the audio where necessary, and add precise conversational tags—specifically turn-taking cues and backchanneling. The content is mostly casual chat, so there is no need to identify speakers, emotions, or specialized domains such as technology, healthcare, or education.
You will work inside our browser-based interface that lets you zoom the waveform, set boundaries, and choose labels from a predefined menu. I will share a short style guide that explains when to mark an utterance as a backchannel (“uh-huh”, “right”) versus a speaker hand-off, plus a set of sample files so you can practice before the real work begins.
Key details
• Applications close in 3 days.
• I need at least 10 hours of work from you before June 7th; more is welcome if you have the time.
• Accuracy is critical—every file you finish will be spot-checked against the guidelines, and any corrections must be applied promptly.
Deliverables
1. Fully segmented audio files with turn-taking and backchanneling tags applied.
2. A brief note on any edge cases or guideline ambiguities you encounter.
If you’re comfortable with English conversational nuances and can start immediately, I’d love to bring you on.
You will work inside our browser-based interface that lets you zoom the waveform, set boundaries, and choose labels from a predefined menu. I will share a short style guide that explains when to mark an utterance as a backchannel (“uh-huh”, “right”) versus a speaker hand-off, plus a set of sample files so you can practice before the real work begins.
Key details
• Applications close in 3 days.
• I need at least 10 hours of work from you before June 7th; more is welcome if you have the time.
• Accuracy is critical—every file you finish will be spot-checked against the guidelines, and any corrections must be applied promptly.
Deliverables
1. Fully segmented audio files with turn-taking and backchanneling tags applied.
2. A brief note on any edge cases or guideline ambiguities you encounter.
If you’re comfortable with English conversational nuances and can start immediately, I’d love to bring you on.