Ezra M.

Key Skills

Software

No software listed

Top Subject Matter

LLM output evaluation and RLHF data feedback

Computer vision and multimodal dataset annotation

Top Data Types

Image

Video

Document

Top Task Types

RLHF

Segmentation

Transcription

Freelancer Overview

Data Annotation Specialist — Outlier AI (Remote). Core strengths include Outlier AI and Atlas Capture. Education includes Diploma in Building and Construction Technology, N/A (2023) and Pedagogy, N/A (2024). AI-training focus includes data types such as Text and Image and labeling workflows including RLHF and Segmentation.

Labeling Experience

Data Annotation Specialist — Atlas Capture (Remote)

ImageSegmentation

Created computer-vision annotations including bounding boxes, keypoints, polygons, and semantic segmentation masks for training datasets. Labeled multimodal data spanning image, audio, and video to support perception and recognition AI systems. Conducted thorough quality checks on peer annotations to maintain dataset integrity and consistency standards. • Applied labeling taxonomies accurately with minimal error rates • Performed precise object annotation for CV model training • Verified annotation quality through peer review • Adapted quickly to new project guidelines and tools

2024 - Present

Data Annotation Specialist — Outlier AI (Remote)

TextRLHF

Provided human feedback on AI-generated text, code, and creative writing to improve large language model (LLM) performance. Performed ranking, rating, and rewriting of model responses to support RLHF pipelines. Assessed outputs for factual accuracy, coherence, tone, and safety across diverse subject domains. • Evaluated model responses for quality and benchmark alignment • Supported team consistency by flagging ambiguous instructions • Contributed to improving annotation guidelines • Maintained high daily throughput while preserving quality scores

2024 - Present

Education

Work History