ML Engineer — EduBridge Adaptive Engine

About the role

The adaptive learning engine is EduBridge's most innovative technical component and differentiator. You will build the machine learning systems that make the platform genuinely adaptive rather than just branching — a system that models each student's knowledge state, predicts which content they are most likely to benefit from next, and automatically adjusts the difficulty and format of assessments. Your primary deliverables: a knowledge tracing model using deep knowledge tracing (DKT) or Bayesian knowledge tracing (BKT) that estimates each student's mastery of each learning objective from their assessment performance history; a next-item recommendation algorithm that selects the optimal learning activity for each student at each moment; a question generation system using LLM APIs that creates novel assessment questions from learning objective definitions and course content; a plagiarism-resistant submission analysis system; and an early warning model that identifies students at risk of dropping out based on engagement and performance signals. You will instrument the platform to collect the learning interaction data (time on task, hint requests, error patterns, replay behaviour) that feeds your models. You will build the feature pipeline, model training infrastructure, and A/B testing framework to evaluate whether the adaptive engine genuinely improves learning outcomes versus a static curriculum. Experience with educational data mining, knowledge tracing models, or learning analytics is a strong advantage. Experience with LLM-based content generation and evaluation is required. This role includes authorship on any research publications.