Multimodal Image Evaluation Annotator

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. OpenTrain AI hires and contracts contributors for projects that help shape how modern AI systems understand and respond to real-world information.

You can create a free OpenTrain account, build a professional AI training profile, and apply in minutes. This role offers an opportunity to develop practical experience in multimodal evaluation, research, image handling, and quality-focused documentation.

About AI Training Work

AI training is the human side of building artificial intelligence. Contributors prepare examples, evaluate model outputs, and create reliable reference materials that help AI systems learn from images, text, audio, and other information.

Multimodal evaluation work focuses on how AI interprets visual information alongside web research and written instructions. Careful human judgment helps make these systems more accurate, useful, and grounded in evidence.

The Role

OpenTrain is recruiting a Multimodal Image Evaluation Annotator to create evaluation materials for AI systems working with visual information and web research. You will work with authentic photographs of real-world places, products, and media while preserving realistic visual clues and concealing key identifying details.

The role combines image sourcing, copyright and licensing verification, investigative research, image editing, prompt writing, rubric development, and documentation of ground-truth answers. The engagement is remote and intended for an entry-level contributor with strong English and careful analytical skills.

• Remote contractor engagement
• Part-time schedule of at least 20 hours per week
• At least 4 hours per day required
• Includes 4 hours of overlap with Pacific Time
• Engagement may last up to 16 weeks
• Compensation follows standard OpenTrain project defaults for this opportunity

What You'll Do

You will source and prepare visual evaluation tasks based on relevant, authentic photographs. Your work should retain realistic secondary details while preventing the image or prompt from directly revealing the target identity or answer.

You will also research incomplete visual clues, document evidence, and create objective materials that allow an AI system's answer to be evaluated consistently.

• Source high-quality photographs for an assigned locale, including landmarks, universities, municipal buildings, products, and domestic film or media.
• Verify whether images are original, publicly available, copyright-free, or explicitly licensed.
• Document each image's provenance and license type.
• Use Google Search, Google Lens, reverse image search, and other reliable sources to identify subjects from partial visual clues.
• Create cropped, occluded, masked, or degraded image versions while retaining realistic secondary details.
• Write time-bound English prompts that do not reveal the target identity or answer.
• Produce concise, objective grading criteria and solution notes.
• Document the research process, evidence, expected outcome, and ground-truth identity.

Requirements

This role requires native or highly fluent English and the ability to communicate clearly in prompts, rubrics, and solution notes. You should be comfortable researching unfamiliar subjects from incomplete visual information and documenting your reasoning in a structured way.

Strong familiarity with culturally significant but non-touristy places, institutions, products, or media in the assigned locale is important. You should also understand practical image licensing, copyright, and provenance verification, and be able to perform basic image edits.

• Native or highly fluent English writing ability
• Investigative web research skills using search and reverse-image tools
• Strong local cultural knowledge relevant to the assigned locale
• Practical understanding of image licensing, copyright, and provenance
• Basic ability to crop, mask, or edit images
• Careful rubric-writing and solution-note skills
• Attention to detail with shared Google Drive folders, file naming, and structured submissions

Who Should Apply

This opportunity may suit people who enjoy visual investigation, online research, cultural discovery, and precise written analysis. Prior AI evaluation, data annotation, content review, quality assurance, or other analytical experience is helpful but not required.

A bachelor's degree or equivalent practical experience in any field is welcome. The listed experience level is entry level, so demonstrated care, strong judgment, and the ability to follow structured requirements are especially valuable.

• Researchers who can identify subjects from partial visual clues
• Writers who can create clear prompts, rubrics, and evidence-based notes
• Contributors familiar with local places, institutions, products, or media
• Detail-oriented workers comfortable organizing files and documenting sources
• Applicants seeking flexible, remote experience in AI training and evaluation

How to Get Started

Create a free OpenTrain account to build your AI training profile and apply for this opportunity in minutes. OpenTrain helps contributors discover and grow through work that spans data labeling, evaluation, research, and other ways people teach AI.

If selected, you will complete the project according to its remote contractor schedule, time-zone overlap requirement, and submission standards. This role can help you build a credible portfolio of multimodal AI evaluation work.

• Language: English
• Work arrangement: Remote and worldwide
• Engagement type: Contractor and part time
• Availability: 20 or more hours per week
• Pacific Time overlap: 4 hours required
• Maximum engagement length: 16 weeks