Project "Theos" (Long Term)

Greetings from Freelancer.com

A new project has come in from one of our tech giant clients, and we are actively looking for native speakers of the following languages

Monolingual - fr-FR, de-DE, pt-BR, hi-IN, ja-JP, ko-KR, es-MX
(French, German, Portuguese, Hindi, Japanese, Korean, Spanish)

Code-mixing combinations - fr-en, de-en, pt-en, hi-en, ja-en, ko-en, ja-ko, es-en
(French, German, Portuguese, Hindi, Japanese, Korean, Spanish)

Project Objectives:
Rater is expected to do testing & rate the conversation with AI model. We require raters who can speak English and their native language to do E2E testing with voice. English Proficiency isn’t important. A native speaker in their languages who knows the basic level of English is sufficient. The target languages are de, fr, id, ja, ko.


• This is a multi-turn test with voice interaction.
• A goal and the first prompt will be provided. The rater should continue conversation with an AI model to achieve the goal.
• Raters will use their mobile phones for testing.
• Raters don’t need to be in-country.
• No need to record the tests with video/camera.
• It’ll be 3 replications.

Eligibility:
• In order to be eligible for this project, you must agree to the Populii terms and conditions as well as the NDA.
• Must be a native resource in languages selected
• Native language proficiency is needed for the evaluation of set language.
• Beginner to intermediate proficiency in English is okay.
• Tasks will be allocated to your account and work availability notifications will be shared on your Populii ID.

Qualifications:
• To work on the Rating Project:
• You must meet the eligibility criteria above
• At least 0-1 years of experience in content rating
• Proficiency in native language, including nuances of grammar
• Understand and comprehend the knowledge base training materials
• Pass the assessment thereafter on Populii with a minimum quality score of 85%
• Maintain an average handling time of 7-15 mins depending on tasks