Key Skills

Software

Appen

Datasaur

Doccano

Figure Eight

Humanatic

Labelbox

LightTag

Lionbridge

Prodigy

Top Subject Matter

Sentiment analysis/Digital interaction data

Identification of dangerous situations/Risk dialogues

LLM evaluation in Brazilian Portuguese

Top Data Types

Image

Text

Video

Top Task Types

Action Recognition

Emotion Recognition

Object Detection

Question Answering

Red Teaming

Freelancer Overview

Tenho experiência prática em rotulagem de dados e treinamento de IA, com foco em análise semântica, avaliação de sentimentos e reconhecimento de padrões contextuais. Ao longo dos projetos, desenvolvi habilidades para identificar e categorizar dados de forma precisa, utilizando técnicas como reconhecimento de ações em vídeos, análise de interações digitais e identificação de situações de risco em diálogos. Essa experiência me permitiu contribuir para a criação de sistemas de IA mais eficientes e alinhados ao comportamento humano. Sou apaixonada por trabalhar com projetos que tornam a IA mais humanizada e responsável, aplicando métodos que garantem a precisão dos modelos e, ao mesmo tempo, promovem impacto social positivo. Minha capacidade de atenção aos detalhes e compromisso com resultados de alta qualidade me diferenciam, especialmente em projetos voltados à personalização e ao aperfeiçoamento de tecnologias para melhorar a experiência do usuário e a segurança em diferentes aplicações.

Labeling Experience

Red Teaming for AI Text-Based Models

AppenTextClassificationQuestion Answering

Participated in a Red Teaming project to test the ethical and functional robustness of AI models handling text-based inputs. My tasks included crafting adversarial scenarios and prompts to expose vulnerabilities in text generation and classification models. Additionally, I annotated responses to flag inappropriate outputs, biases, and inconsistencies, ensuring the model adhered to ethical standards and improved reliability.

Participated in a Red Teaming project to test the ethical and functional robustness of AI models handling text-based inputs. My tasks included crafting adversarial scenarios and prompts to expose vulnerabilities in text generation and classification models. Additionally, I annotated responses to flag inappropriate outputs, biases, and inconsistencies, ensuring the model adhered to ethical standards and improved reliability.

2024 - 2024

Sentiment annotation in digital interactions

ProdigyTextEntity Ner ClassificationQuestion Answering

I collaborated on a project focused on sentiment analysis, where I was responsible for identifying emotions and intentions in texts, such as customer support messages. My role involved classifying emotional patterns (e.g., anxiety, frustration, empathy) and detailed annotation of dialogue intentions. The goal was to help the AI provide more humanized and ethical responses, especially in sensitive contexts like mental health support.

I collaborated on a project focused on sentiment analysis, where I was responsible for identifying emotions and intentions in texts, such as customer support messages. My role involved classifying emotional patterns (e.g., anxiety, frustration, empathy) and detailed annotation of dialogue intentions. The goal was to help the AI provide more humanized and ethical responses, especially in sensitive contexts like mental health support.

2024 - 2024

Education

Work History