Conversational Speech AI Evaluator; United States (Spanish)

About the role

Project Astra is seeking individuals for an ongoing, ad-hoc project focused on improving real-time conversational AI. In this project, participants will converse with a speech-to-speech AI model and evaluate its performance across speech recognition, audio output quality, conversation flow, and content accuracy. Help improve real-time conversational AI by evaluating multilingual voice interactions for accuracy, quality, and natural flow.

Responsibilities

• Engage in conversations with a real-time speech-to-speech AI model
• Evaluate performance based on speech recognition, audio quality, conversation flow, and content accuracy
• Provide accurate and consistent ratings based on project guidelines
• Complete assigned tasks within the given timelines

Each conversation will last approximately 8 to 10 minutes.

Requirements

• Must be a native Spanish speaker residing in the United States
• Must have a strong understanding of grammar and localization
• Keen attention to detail
• Access to a stable internet connection and a computer

Pay

$21 - $21 an hour