Helena Costa

Key Skills

Software

Appen

Data Annotation Tech

Remotasks

Scale AI

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Image

Top Task Types

Action Recognition

Audio Recording

Classification

Emotion Recognition

Evaluation Rating

Freelancer Overview

With a strong background in digital strategy, communication, and project management, I bring over twenty years of experience in detail-oriented, high-judgment work to my data annotation and AI training projects. I’ve contributed to a wide range of annotation projects, either in English, Portuguese or Spanish, including general data labeling, video labeling, and projects focused on user safety and model safety. Across all assignments, I’ve maintained a commitment to precision, clarity, and ethical decision-making, ensuring that each dataset supports responsible and high-quality model development. My experience has strengthened my ability to interpret complex guidelines, assess nuanced content, and apply consistent judgment even in ambiguous cases. I approach annotation with a balance of analytical thinking and contextual understanding, aiming to enhance model reliability and fairness. I’m also skilled in quality review and cross-functional collaboration, helping refine processes and uphold data integrity standards. I value OpenAI’s focus on safety and alignment, and I take pride in contributing to systems that reflect human values and accuracy. My goal is to continue supporting the creation of trustworthy, well-trained models through careful data labeling and thoughtful evaluation.

Labeling Experience

AI Data Annotation & Safety Labeling Specialist

Data Annotation TechDocumentBounding BoxClassification

I contributed to a range of data annotation projects supporting AI model training and safety evaluation in three different languages (Portuguese, English and Spanish). Tasks included text, audio, image, and video labeling; content classification; and context-based judgment for user and model safety initiatives. I applied detailed annotation guidelines to ensure accuracy, consistency, and ethical data interpretation across diverse content types. Projects often required identifying nuanced context, assessing potential safety risks, and evaluating model outputs for alignment and reliability. I collaborated with cross-functional QA teams to maintain quality standards, provided feedback on ambiguous cases, and consistently achieved high accuracy and throughput rates. My work directly contributed to improving model performance, safety, and contextual understanding in real-world applications.

I contributed to a range of data annotation projects supporting AI model training and safety evaluation in three different languages (Portuguese, English and Spanish). Tasks included text, audio, image, and video labeling; content classification; and context-based judgment for user and model safety initiatives. I applied detailed annotation guidelines to ensure accuracy, consistency, and ethical data interpretation across diverse content types. Projects often required identifying nuanced context, assessing potential safety risks, and evaluating model outputs for alignment and reliability. I collaborated with cross-functional QA teams to maintain quality standards, provided feedback on ambiguous cases, and consistently achieved high accuracy and throughput rates. My work directly contributed to improving model performance, safety, and contextual understanding in real-world applications.

2025

Education

Work History