Claudia Konar

Key Skills

Software

AWS SageMaker

Appen

Clickworker

Mindrift

Remotasks

Scale AI

Surge AI

Toloka

Labelbox

Top Subject Matter

No subject matter listed

Top Data Types

Document

Image

Text

Top Task Types

Classification

Emotion Recognition

Evaluation Rating

Prompt Response Writing SFT

Text Summarization

Freelancer Overview

As an AI training contributor with a strong background in psychology and human behavior, I specialize in evaluating and annotating emotionally nuanced content. My work includes text classification, NER (Named Entity Recognition), RLHF (Reinforcement Learning from Human Feedback), and prompt + response alignment across NLP projects. I’ve contributed to improving large language models through hands-on annotation tasks involving ethical reasoning, emotion recognition, and psychological intent analysis. My interdisciplinary knowledge and attention to linguistic subtleties help me generate high-quality data for ethical and human-aligned AI.

Labeling Experience

Video Classification & Annotation

Surge AIVideoClassification

Watched and annotated short-form videos to classify behavioral context, emotional tone, and safety concerns. Labeled scenes for actions, sentiment, and compliance with content policies. Applied human judgment in detecting sarcasm, implicit meaning, and potential psychological impact. Supported development of AI models for social media moderation and behavioral pattern detection.

Watched and annotated short-form videos to classify behavioral context, emotional tone, and safety concerns. Labeled scenes for actions, sentiment, and compliance with content policies. Applied human judgment in detecting sarcasm, implicit meaning, and potential psychological impact. Supported development of AI models for social media moderation and behavioral pattern detection.

2025

Product Review Sentiment Analysis

Scale AITextClassificationEmotion Recognition

Labeled e-commerce product reviews for sentiment (positive, neutral, negative) and customer intent. Helped train models to understand sarcasm, dissatisfaction, excitement, and subjective feedback across product categories. Improved AI-driven sentiment analysis systems in retail and marketing contexts.

Labeled e-commerce product reviews for sentiment (positive, neutral, negative) and customer intent. Helped train models to understand sarcasm, dissatisfaction, excitement, and subjective feedback across product categories. Improved AI-driven sentiment analysis systems in retail and marketing contexts.

2025

AI Ethics Filtering

Scale AITextClassification

Reviewed AI-generated outputs and tagged potentially unethical, manipulative, or privacy-violating content. Applied context-based ethical reasoning to classify responses and support training of safety-first AI models. Maintained high precision in judgment-based annotation tasks.

Reviewed AI-generated outputs and tagged potentially unethical, manipulative, or privacy-violating content. Applied context-based ethical reasoning to classify responses and support training of safety-first AI models. Maintained high precision in judgment-based annotation tasks.

2025

Social Media Content Moderation

AppenImageClassification

Labeled user-generated content from social media platforms for potential policy violations. Tagged offensive language, misinformation, hate speech, and unsafe imagery. Ensured moderation aligned with platform guidelines and international safety standards.

Labeled user-generated content from social media platforms for potential policy violations. Tagged offensive language, misinformation, hate speech, and unsafe imagery. Ensured moderation aligned with platform guidelines and international safety standards.

2025

Classification

LabelboxTextClassification

Classified text content based on educational level (K-12, college, professional), subject matter, and instructional intent. Supported AI models in tailoring responses to appropriate cognitive levels and didactic goals. Ensured contextual accuracy in learning materials across multiple disciplines.

Classified text content based on educational level (K-12, college, professional), subject matter, and instructional intent. Supported AI models in tailoring responses to appropriate cognitive levels and didactic goals. Ensured contextual accuracy in learning materials across multiple disciplines.

2025

Show more

Education

Work History