Anquinette J.

Key Skills

Software

Label Studio

Top Subject Matter

No subject matter listed

Top Data Types

Text

Video

Image

Top Task Types

RLHF

Entity (NER) Classification

Transcription

Freelancer Overview

As a detail-oriented professional with a background in data evaluation and quality assurance, I specialize in enhancing AI performance through precise data annotation and objective output assessment. My experience includes analyzing complex prompts, identifying subtle logical inconsistencies, and ensuring generated content aligns with strict safety and quality benchmarks. I pride myself on maintaining high inter-rater reliability and producing consistent, high-quality labels that directly improve model training iterations. Beyond fundamental labeling, I bring subject matter expertise that enables me to handle specialized datasets, such as technical, creative, or reasoning-based tasks. By combining strong analytical thinking with an efficient workflow, I consistently meet project deadlines without sacrificing accuracy. Whether evaluating long-form content or performing fine-grained classification, I am committed to providing the nuanced feedback necessary to build safer, more reliable, and more helpful AI systems.

Labeling Experience

Natural Language Processing (NLP) Annotation

TextEvaluation/Rating

I served as a lead data annotator for a large-scale project aimed at improving the natural language reasoning capabilities of a generative AI model. My primary task was multi-class text classification and output evaluation, where I assessed AI-generated responses for factual accuracy, adherence to safety guidelines, and logical consistency. The scope of this project involved processing high volumes of unstructured conversational data to refine the model's ability to handle nuanced user queries. Over the course of this engagement, I contributed to a dataset comprising over 5,000 unique interactions. To ensure data integrity, I strictly adhered to an evolving project rubric that required 95% inter-rater reliability. I performed regular peer-review audits to minimize labeling bias and consistently met tight weekly output quotas while maintaining a high quality-assurance score, directly contributing to the model's iterative improvement in accuracy and tone.

2020 - 2021

Education

Work History