Romir Rajendra M.

Key Skills

Software

SuperAnnotate

Mercor

Other

Top Subject Matter

Visual and multimedia AI evaluation (NVIDIA project)

Multimodal AI model evaluation and training

Legal Services & Contract Review

Top Data Types

Image

Text

Document

Top Task Types

Segmentation

Evaluation/Rating

Prompt + Response Writing (SFT)

Transcription

RLHF

Classification

Bounding Box

Object Detection

Text Generation

Question Answering

Freelancer Overview

AI Trainer — Prompt & Output Evaluator. Brings 6+ years of professional experience across complex professional workflows, research, and quality-focused execution. Core strengths include SuperAnnotate and Outlier Platform. Education includes Master of Science, University of Liverpool (2025) and Bachelor of Science, Nagpur University (2022). AI-training focus includes data types such as Image and Text and labeling workflows including Evaluation and Rating.

Labeling Experience

AI Trainer — Prompt & Output Evaluator

SuperAnnotateImage

As an AI Trainer for SuperAnnotate’s NVIDIA project, I evaluated AI-generated text and visual outputs for quality, relevance, and safety. My work involved structured review and rating of large datasets, supporting model development with performance benchmarks and quality scores. I identified strengths and flagged inconsistencies in outputs, demonstrating objective, guideline-driven analysis. • Assigned structured ratings and justifications for both prompt and image-based tasks. • Flagged and documented model failure modes with actionable written feedback. • Worked independently in a high-volume, remote setting. • Utilized the SuperAnnotate platform throughout all tasks.

2026 - Present

AI Model Trainer

Text

As an AI Model Trainer at Outlier, I conducted structured AI evaluation across text, image, video, audio, and website generation models. I provided actionable feedback and rankings of AI outputs based on guideline compliance, relevance, and accuracy. I also tested semantic alignment and quality for various media types via annotation platforms. • Validated audio transcripts for accuracy. • Ranked and reviewed image, video, and website outputs. • Utilized Python and JSON for reviewing and organizing data. • Leveraged Outlier’s digital annotation tools for all task submissions.

2025 - Present

Education

Work History