Romanian Audio Generalist Evaluator Expert (SF or London, in person)
Mercor • Remote • Posted 22 days ago
Education
Any
Type
Pay Rate
$37.5/task
Posted
22d ago
✅ Applying through this link gives you a verified candidate referral.
Referrals from verified candidates give your profile a visibility boost and help support our platform at no cost to you.
This position is hosted on an external talent platform. Please only apply for this position if it fits your skills and interests.
About this Role
Mercor is seeking detail-oriented transcribing/writing experts to contribute to a high-impact audio AI research project with a leading lab. Freelancers will listen to and transcribe audios, annotate images and evaluate videos to help train advanced language models. This is a ~6 month short-term, flexible opportunity for professionals with strong academic backgrounds and a knack for instructional clarity. Ideal for those who enjoy distilling complex concepts into well-crafted text.
Transcribe and Optimize Audio/Video: Create detailed audio transcriptions with multiple constraints and instructions from Romanian.
Location Requirements
Job Details
Transcribe and Optimize Audio/Video
Define and Document Evaluation Standards
Conduct Model Testing and Grading
Support Benchmarking and Quality Assurance
Minimum Qualifications
Preferred Qualifications
More Details About This Role
20 hours per week
Mercor is seeking detail-oriented transcribing/writing experts to contribute to a high-impact audio AI research project with a leading lab. Freelancers will listen to and transcribe audios, annotate images and evaluate videos to help train advanced language models. This is a ~6 month short-term, flexible opportunity for professionals with strong academic backgrounds and a knack for instructional clarity. Ideal for those who enjoy distilling complex concepts into well-crafted text. Transcribe and Optimize Audio/Video: Create detailed audio transcriptions with multiple constraints and instructions from Romanian. Define and Document Evaluation Standards: Establish high-level expectations for correct responses in general audio consumer contexts, and develop comprehensive rubric. Conduct Model Testing and Grading: Run prompts through models and assess preliminary outputs against expectations. Support Benchmarking and Quality Assurance: Collaborate in QA review processes to ensure prompt tasks and rubrics meet rigor, maintaining consistency and reliability before integration into official benchmarks. BS or BA from a reputable institution completed or in progress Strong writing and critical thinking skills. Complete an AI-led interview, this should take around 15 minutes. If selected, you will be invited to work on the project. Expect to contribute at least 20 hours per week. Expect a commitment of around 2-4 weeks. We consider all qualified applicants without regard to legally protected characteristics and provide reasonable accommodations upon request.
- Transcribe and Optimize Audio/Video: Create detailed audio transcriptions with multiple constraints and instructions from Romanian.
- Define and Document Evaluation Standards: Establish high-level expectations for correct responses in general audio consumer contexts, and develop comprehensive rubric.
- Conduct Model Testing and Grading: Run prompts through models and assess preliminary outputs against expectations.
- Support Benchmarking and Quality Assurance: Collaborate in QA review processes to ensure prompt tasks and rubrics meet rigor, maintaining consistency and reliability before integration into official benchmarks.
- BS or BA from a reputable institution completed or in progress
- Strong writing and critical thinking skills.
- College students/grads
- Complete an AI-led interview, this should take around 15 minutes.
- If selected, you will be invited to work on the project.
- Expect to contribute at least 20 hours per week.
- Expect a commitment of around 2-4 weeks.
Requirements
- Must be eligible to work in Remote
- Fluent proficiency in English (Written & Verbal)
- Reliable high-speed internet connection
Eligible Languages
Fluent proficiency in Romanian
Compensation Analysis
Rare opportunity for top 1% experts. Earn $37.5/hr contributing to the world's most advanced AI labs. This is one of the few roles where academic precision is valued as highly as commercial output.
Skills & Categories
Explore other opportunities in related specializations:
Related Jobs
Browse All Jobs from Mercor
Discover more opportunities on Mercor that match your skills and interests.
View All Mercor Jobs →Community Reviews
Leave your review
Frequently Asked Questions
Is this for freelancers or full-time employees?
Both. Mercor tries to match you with clients who want long-term contractors. Unlike other platforms where you log in and grab small tasks, Mercor matches you with one company for a steady role (e.g., 'Python Tutor for 3 months').
I'm not comfortable on camera. Can I still apply?
No. The application requires a video interview with an AI avatar. The AI asks you questions about your resume, and the video is shared with potential clients to prove your communication skills.
Does it cost money to join?
No. You should never pay to join these platforms. Mercor makes money by charging the client a fee on top of your hourly rate.
What equipment do I need?
For voice or audio roles at this pay level, you typically need a professional home studio setup (XLR microphone, treated room). Phone recordings or laptop mics are usually rejected by quality control.
How is my work used?
You are providing high-quality "ground truth" data. For writers, this means creative generation. For voice actors, it often means training Text-to-Speech models. Be sure to check the specific contract details regarding rights usage for your voice or likeness.
Is creative freedom allowed?
Yes and no. While you are hired for your talent, you must often follow strict style guides (e.g., "Speak in a neutral tone" or "Write in the style of a technical manual"). The goal is consistency for the dataset.
How soon will I start?
Important: Mercor is a talent marketplace, not a task queue. Applying puts you in a pool of candidates. You will only start working when a specific client (like a major AI lab) selects your profile. This matching process can take weeks.