Lead Data Engineer / Researcher — Student Attrition Predictive Academic Data Project (Framework Present, 2025–Present)
You led a student attrition predictive data framework by creating training-ready labeled data from raw academic grades and engagement metadata. Your work included taxonomy protocol creation, manual feature labeling, and cleaning incomplete or noisy records. You ensured label reliability by auditing output distributions to protect against algorithmic bias and maintain pristine ground-truth labels. • Data curation and target-matrix modeling for labeled training data • Cleaning & preprocessing of incomplete datasets and noise profiles • Manual labeling of behavioral features to secure accurate training parameters • Validation via auditing output distributions to reduce bias