George Tanios

Key Skills

Software

Scale AI

Other

Top Subject Matter

No subject matter listed

Top Data Types

Audio

Document

Image

Text

Video

Top Task Types

RLHF

Data Collection

Prompt Response Writing SFT

Audio Recording

Transcription

Freelancer Overview

I specialize in AI model evaluation and data annotation, with hands-on experience improving large language model (LLM) performance through structured response ranking, detailed failure analysis, and dataset refinement. My work at Outlier AI involved evaluating and ranking hundreds of LLM outputs across reasoning, summarization, and knowledge-based tasks, where I applied rigorous scoring rubrics and identified issues such as hallucinations, logical inconsistencies, and safety risks. I am skilled in using tools like Google Sheets and Excel for QA tracking and structured scoring, and I have a strong understanding of annotation taxonomies, error categorization, and adversarial prompt testing. My ability to deliver high-quality, consistent results under strict QA standards makes me confident in handling complex NLP and AI training data projects.

Labeling Experience

AI Trainer

Don T DiscloseImageRLHFData Collection

Evaluated and ranked 500+ LLM-generated responses across reasoning, summarization, and knowledge-based tasks Applied structured scoring rubrics assessing coherence, factuality, alignment, and instruction adherence Identified hallucinations and categorized failure modes (fabrication, logical gaps, unsupported claims) Maintained 95%+ task acceptance rate under QA audits Delivered written rationales to support comparative ranking decisions Contributed to refinement of training datasets used to improve model performance Completed time-sensitive batches while meeting quality benchmarks and consistency standards

Evaluated and ranked 500+ LLM-generated responses across reasoning, summarization, and knowledge-based tasks Applied structured scoring rubrics assessing coherence, factuality, alignment, and instruction adherence Identified hallucinations and categorized failure modes (fabrication, logical gaps, unsupported claims) Maintained 95%+ task acceptance rate under QA audits Delivered written rationales to support comparative ranking decisions Contributed to refinement of training datasets used to improve model performance Completed time-sensitive batches while meeting quality benchmarks and consistency standards

2024 - 2025

AI Trainer

Scale AIAudioText GenerationEmotion Recognition

Responsible for reviewing and annotating audio recordings to support the development of speech recognition and machine learning systems. Listened to diverse audio clips and accurately transcribed spoken content while identifying speakers, background noises, tone, and other relevant sound characteristics. Applied detailed labeling guidelines to ensure consistency and high data quality. Maintained accuracy under time constraints, flagged unclear audio segments when necessary, and contributed to improving AI model performance through precise and reliable annotations. Demonstrated strong attention to detail, critical listening skills, and the ability to follow complex instructions independently.

Responsible for reviewing and annotating audio recordings to support the development of speech recognition and machine learning systems. Listened to diverse audio clips and accurately transcribed spoken content while identifying speakers, background noises, tone, and other relevant sound characteristics. Applied detailed labeling guidelines to ensure consistency and high data quality. Maintained accuracy under time constraints, flagged unclear audio segments when necessary, and contributed to improving AI model performance through precise and reliable annotations. Demonstrated strong attention to detail, critical listening skills, and the ability to follow complex instructions independently.

2024

Education

Work History