Bulgarian-Speaking AI Language Evaluator

You will help improve language models by reviewing AI-generated Bulgarian text, creating reference answers, and evaluating responses against clear quality standards. Your work will cover general knowledge, customer support, education, conversational writing, translation, localization, and culturally sensitive content.

• Create Bulgarian prompts and high-quality reference answers.
• Review grammar, spelling, fluency, tone, cultural fit, terminology, and logical consistency.
• Check whether responses follow the prompt and preserve the intended meaning.
• Write detailed feedback, accurate rewrites, and model answers in clear, natural Bulgarian.
• Compare and rank AI responses for correctness, fluency, fidelity, register, safety, terminology, and overall quality.
• Identify mistranslations, unsupported claims, bias, unsafe language, awkward literal phrasing, and culturally mismatched wording.

What it pays and takes

This is a remote, part-time contractor role for freelancers located in Bulgaria. You will work independently with flexibility across time zones.

• Pay: Up to $20 per hour.
• Time requirement: 20 or more hours per week.
• Location: Bulgaria only.
• Languages: Fluent Bulgarian and professional English at C1 level or higher.
• Writing knowledge: Strong command of Bulgarian grammar, spelling, punctuation, Cyrillic conventions, tone, style, and different registers.
• Experience: At least three years in writing, editing, translation, localization, content quality, journalism, education, or related language work.
• Education: A bachelor's degree or higher in a relevant language, writing, communications, or education field, or equivalent professional experience.
• You must be able to explain language errors and suggest corrections that preserve meaning and suit the intended audience.

How it works

Apply on OpenTrain with your resume, then complete the application on the hiring site.

About AI training work

AI training work uses human judgment to teach and evaluate systems that generate language and other content. People with strong language skills are needed to identify errors, explain what good responses look like, and help models produce more accurate and useful results.