Felix Abonchez

Key Skills

Software

iMerit

Redbrick AI

Scale AI

Surge AI

Top Subject Matter

No subject matter listed

Top Data Types

Image

Text

Top Label Types

Text Generation

Text Summarization

RLHF

Data Collection

Freelancer Overview

I am an experienced AI and NLP specialist with a strong background in data labeling, content evaluation, and training data quality for machine learning systems. My work has focused on developing and leading frameworks for assessing and annotating AI-generated text, ensuring accuracy, coherence, and contextual relevance at scale. I have hands-on expertise with Python, Java, and TypeScript, as well as advanced knowledge of ML frameworks such as TensorFlow, PyTorch, and Hugging Face Transformers. My experience spans leading teams in building evaluation protocols, automating annotation workflows, and publishing research on quality assessment metrics for large language models. I am passionate about improving the reliability of AI through robust data annotation and have contributed to both open-source NLP libraries and patented content evaluation methodologies.

Labeling Experience

Large-Scale Evaluation and Annotation of AI-Generated Text for Quality and Safety

Scale AITextText GenerationText Summarization

Conducted large-scale text data labeling and evaluation projects focused on AI-generated content quality, safety, and alignment. Responsibilities included annotating and rating model outputs for coherence, factual accuracy, contextual relevance, bias, and harmful content across multiple NLP tasks such as question answering, summarization, and prompt–response generation. Designed and applied detailed annotation guidelines, performed multi-pass quality checks, and contributed to RLHF pipelines by generating high-quality prompts and reference responses. Project scale exceeded millions of text samples, with strict inter-annotator agreement thresholds and automated validation workflows to ensure consistency and reliability.

Conducted large-scale text data labeling and evaluation projects focused on AI-generated content quality, safety, and alignment. Responsibilities included annotating and rating model outputs for coherence, factual accuracy, contextual relevance, bias, and harmful content across multiple NLP tasks such as question answering, summarization, and prompt–response generation. Designed and applied detailed annotation guidelines, performed multi-pass quality checks, and contributed to RLHF pipelines by generating high-quality prompts and reference responses. Project scale exceeded millions of text samples, with strict inter-annotator agreement thresholds and automated validation workflows to ensure consistency and reliability.

2020 - 2023

Education

Work History