Apna
Odia - AI Voice Data Collection Project
Job description
About the Role
We are looking for fluent Odia speakers to join a remote, one-time voice recording project that helps build next-generation AI models.
About the Project
We are looking for fluent Odia speakers to participate in a conversational voice recording project aimed at training and improving AI speech and conversational models.
Participants will have natural, unscripted conversations with an AI-powered conversational bot across topics such as Travel, Finance, Retail, Healthcare, Telecom, Education, and more.
This is not a script-reading project. We are looking for participants who can speak naturally, confidently, and spontaneously while providing detailed responses.
Selection Process
Audition: Submit a 2-minute voice recording in a quiet environment.
Screening: Submitted samples will be reviewed based on language fluency, pronunciation, voice quality, and conversational ability.
Shortlisting: Selected candidates will be invited to participate in the final recording.
NDA: Selected candidates must sign an NDA before the final recording session.
Compensations: Rs. 500/Hr | Max Rs. 1000 for 2 hours of recording
Requirements
What You'll Do
Record a 2-minute audition sample as part of the selection process.
If selected, complete 2 hours of conversational voice recording from home.
Interact naturally with an AI bot that will guide the conversation through different topics.
Speak for approximately 70% of the overall conversation, providing detailed, spontaneous, and relevant responses.
Participate in multiple short recording sessions, typically around 10 minutes per topic.
Follow the provided guidelines on:
Natural speaking style
Pronunciation and clarity
Conversational flow
Appropriate response length and detail
Re-record sections if required based on quality or technical checks.
Candidate Requirements
Fluent in Odia ; native or near-native proficiency preferred.
Strong ability to speak naturally, clearly, and confidently in the target language.
Comfortable engaging in conversations on a variety of everyday and professional topics.
Able to provide detailed responses without relying on a script.
Prior experience in voice recording, voice-over, speech data collection, or similar projects is preferred but not mandatory.
Access to a quiet and controlled recording environment at home.
Stable and reliable internet connection.
Access to a good-quality microphone or earphones/headset with a microphone.
Ability to follow recording instructions and complete the assigned sessions within the specified timeline.
Important
This is a conversational speech collection project, not a read-aloud or scripted recording task. Candidates should be comfortable speaking spontaneously and maintaining a natural conversation with an AI bot.