Stefano F.

Key Skills

Software

Scale AI

Labelbox

Remotasks

Data Annotation Tech

Top Subject Matter

Technology – Software Engineering & Programming

Cybersecurity – Information Security & Risk Management

Artificial Intelligence – LLM Evaluation & Data Annotation

Top Data Types

Image

Text

Computer Code Programming

Top Task Types

Classification

RLHF

Computer Programming/Coding

Evaluation/Rating

Question Answering

Freelancer Overview

I am an advanced Computer Engineering student and IT Risk professional at EY with hands-on experience in AI training, data annotation, and quality evaluation projects. I have worked on tasks involving response assessment, prompt evaluation, content review, ranking AI-generated outputs, and reinforcement learning from human feedback (RLHF). My work requires applying detailed guidelines to evaluate accuracy, reasoning quality, instruction-following, safety, and overall response effectiveness. My background in software engineering, cybersecurity, databases, and systems analysis allows me to evaluate both technical and non-technical content with precision. I am highly detail-oriented, analytical, and comfortable working with large-scale annotation projects that require consistency, quality assurance, and critical thinking. My combination of technical expertise and AI evaluation experience enables me to contribute effectively to data labeling, model training, and AI improvement initiatives.

Labeling Experience

AI Response Evaluation & RLHF Specialist

TextClassification

Worked on AI training and evaluation projects involving reinforcement learning from human feedback (RLHF), response ranking, prompt analysis, and quality assessment of AI-generated content. Evaluated model outputs for factual accuracy, reasoning quality, instruction adherence, safety, and overall effectiveness according to detailed annotation guidelines. Performed comparative assessments between multiple AI responses, identified hallucinations and logical inconsistencies, and provided structured feedback to improve model performance. Worked with multilingual content (English and Spanish) across technical and non-technical domains while maintaining high quality standards, consistency, and attention to detail.

2026 - Present

Education

Work History