Leaky Barak

Evaluates large language models and AI systems across reasoning, factual accuracy, instruction following, and response quality. Designs evaluation frameworks, annotation guidelines, and scoring rubrics, and performs benchmark testing and error analysis.

Developed Python-based workflows for preparing, processing, validating, and analyzing machine learning datasets. Built data pipelines, conducted model comparison experiments using metrics like accuracy, precision, recall, and F1-score, and applied statistical analysis to identify model strengths and weaknesses.

Evaluated AI-generated responses and machine learning outputs using structured quality criteria. Designed and maintained benchmark tasks, performed qualitative and quantitative analysis, and created annotation guidelines to improve evaluation consistency.

Evaluated AI-generated outputs for accuracy, relevance, and quality. Reviewed and annotated large-scale datasets, performed data quality reviews, and contributed feedback to improve annotation workflows and labeling guidelines.

Completed a comprehensive undergraduate program focused on artificial intelligence, machine learning, computational systems, and software development.

Completed the Machine Learning Specialization, covering supervised and unsupervised learning, model evaluation, and validation methods.

Completed the Generative AI with Large Language Models course, covering LLM fundamentals, prompt engineering, and evaluation techniques.