Alan Rodolfo Moreno

Key Skills

Software

Labelbox

Mercor

Mindrift

OneForma

Remotasks

SuperAnnotate

Top Subject Matter

No subject matter listed

Top Data Types

Computer Code Programming

Image

Text

Top Label Types

RLHF

Evaluation Rating

Computer Programming Coding

Classification

Prompt Response Writing SFT

Freelancer Overview

With over a decade of experience in software engineering and AI research, I specialize in verifying data quality and ensuring the correctness of AI training data across large-scale, multi-language codebases. My expertise lies in rapidly understanding complex systems, identifying edge cases, and meticulously auditing data pipelines for behavioral accuracy—skills that are essential for high-quality data labeling and annotation. I am highly proficient in Python, Go, C++, and JavaScript, and have hands-on experience with tools such as PyTorch, Docker, and advanced CI/CD pipelines. My background includes leading system verification efforts for distributed AI systems, producing clear technical documentation, and collaborating with research teams to uphold data integrity standards in fields like reinforcement learning and AI safety. I thrive in autonomous, project-driven environments and am dedicated to delivering structured, reliable training data that accelerates AI development.

Labeling Experience

Senior AI Research Scientist

MercorImageRLHFEvaluation Rating

As Senior AI Research Scientist, led a team to develop high-quality datasets for a multi-modal model. Focused on image-to-text consistency and logical reasoning. Designed an evaluation axis for technical craftsmanship. Developed an adversarial prompting suite reducing model toxicity by 40%. • Lead a team of 15 trainers. • Designed evaluation axis to identify artifacts. • Developed a suite for adversarial prompting. • Reduced model toxicity by 40%.

As Senior AI Research Scientist, led a team to develop high-quality datasets for a multi-modal model. Focused on image-to-text consistency and logical reasoning. Designed an evaluation axis for technical craftsmanship. Developed an adversarial prompting suite reducing model toxicity by 40%. • Lead a team of 15 trainers. • Designed evaluation axis to identify artifacts. • Developed a suite for adversarial prompting. • Reduced model toxicity by 40%.

2022

AI Training Specialist & Researcher

LabelboxImageClassificationRLHF

As AI Training Specialist, performed rigorous QA on edited images. Ensured unrequested regions remained faithful, meeting high-quality standards. Identified subtle AI markers like warped anatomy and glossy textures. Ensured models met both explicit and implicit requests without contradictions. • Rigorous quality assurance on images. • Maintained original quality in unrequested regions. • Identified subtle AI markers. • Ensured model request compliance.

As AI Training Specialist, performed rigorous QA on edited images. Ensured unrequested regions remained faithful, meeting high-quality standards. Identified subtle AI markers like warped anatomy and glossy textures. Ensured models met both explicit and implicit requests without contradictions. • Rigorous quality assurance on images. • Maintained original quality in unrequested regions. • Identified subtle AI markers. • Ensured model request compliance.

2018 - 2021

Education

Work History