Algorithmic Pedagogy at Scale: Architecture, Automation, Ethics, and Academic Labour in AI-Driven Higher Education

This paper introduces kognita, a comprehensive software platform designed to augment the capabilities of key stakeholders within the (higher) education ecosystem through deeply integrated artificial intelligence. Built on a modern technology stack comprising Next.js, Firebase, and Google's Genkit for AI orchestration, the platform provides a suite of role-specific tools powered by large language models that range from personalised student study plan generation and automated exam marking to AI-agent-led crisis management for entire courses. By integrating generative AI into the specific workflows of students, educators, examiners, and administrators, kognita serves as both a proof-of-concept and a cautionary exploration of a new paradigm in educational technology. This paradigm moves beyond single-function AI tools toward holistic, context-aware systems that promise to enhance pedagogical effectiveness while simultaneously raising profound questions about the future of human expertise in teaching, the ethics of algorithmic instruction, and the socioeconomic implications of AI-driven educational automation. As we stand at the precipice of a transformation that could fundamentally alter the nature of higher education, this platform and the discourse it enables become essential to understanding not merely what AI can do for education, but what education must protect from AI.