Social Media Content Evaluator

About the role

You will stand at the intersection of artificial intelligence and media analysis, directly influencing how systems interpret short-form video content. Your primary focus will be the meticulous review of video media, where you dissect and classify every linguistic element present in the spoken dialogue and the text visible on the screen. This position requires you to operate within your native locale, producing high-quality "ground truth" data that serves as the definitive benchmark for model training. You are the final authority on whether the language and cultural context match the intended target, ensuring authenticity in every frame. This is a freelance engagement designed for individuals who require a flexible daily schedule while committing to a fixed number of productive hours. You will work entirely from your location in India, using your own equipment to access proprietary classification tools. Immediate availability is essential, as the project requires swift integration to meet evolving data demands. Your work will contribute directly to the creation of robust datasets that power the next generation of automated media understanding.

Key facts

What you'll do

Analyze short-form video media to identify and document every language present in the spoken audio track and any on-screen text.

Validate that the media content aligns precisely with your assigned target locale, assessing both linguistic accuracy and cultural relevance.

Generate high-quality ground truth labels by utilizing internal classification tools and adhering strictly to project-specific guidelines.

Ensure the integrity of datasets by applying rigorous quality assurance checks, maintaining consistency necessary for training automated machine learning models.

Classify linguistic components with precision, distinguishing between primary and secondary languages to support complex multilingual analysis.

Verify the authenticity of spoken language within the video, confirming that dialogue matches the expected native or target tongue.

Audit on-screen text to ensure that any visible language is correctly identified and transcribed according to established protocols.

Monitor your daily progress to ensure you meet the required four-hour contribution window while maintaining high standards of accuracy.

Document any anomalies or edge cases in the media that require clarification or further review by the project team.

Support the scalability of data collection by providing reliable and repeatable evaluations that can be replicated across large video libraries.

Contribute to the refinement of classification taxonomies by offering feedback based on the challenges observed during the evaluation process.

Perform detailed comparisons between your assessments and quality checks to identify discrepancies and improve long-term reliability.

Engage with the task interface to input data efficiently, minimizing downtime between video reviews and label submissions.

Act as a critical validator of cultural context, ensuring that translations or language usage do not appear forced or inappropriate for the target audience.

Requirements

Must be currently living in India, as the role requires validation from a local, native perspective.

Must possess native-level written fluency in Tamil, demonstrating an understanding of formal and informal written registers.

Must possess native-level spoken fluency in Tamil, capable of understanding diverse accents and conversational styles.

Must have access to a reliable, high-speed internet connection capable of supporting continuous video streaming and data uploads.

Must own a secure personal computer or laptop that meets the technical specifications of the evaluation platform.

Must be able to strictly follow detailed photography and data entry guidelines without deviation or error.

Must be available to commit to a minimum of four hours of work per day on a consistent basis.

Must be able to start immediately and sustain the required daily schedule for the duration of the contract.

Must have the legal right to work in India and comply with all local tax and contractor regulations.

Must demonstrate a strong attention to detail, ensuring that language classifications are accurate to the smallest textual element.

Must be comfortable working independently for extended periods, managing time effectively without direct supervision.

Must have the ability to distinguish between similar-sounding phonemes and script characters specific to the Tamil language.

Must be willing to adhere to the standardized protocols of data labeling and quality assurance throughout the engagement.

Must be able to recognize and differentiate between multiple languages when they appear concurrently in a single video frame.

Nice to have

Previous experience in data annotation, content moderation, translation, transcription, or linguistic QA is a strong asset.

Practical notes

Work Schedule: Part-time - 4 hours per day required. (Flexible schedule: you can choose your hours as long as the daily time requirement is met). Work whenever you want.

Start Date: Immediately

Duration: TBC

Rate: 6 USD per hour

Location: India (work from home)

Job Type: Freelance