Instructor-Grounded Agentic Ai for Continuous Learning and Career Alignment

The widespread adoption of large language models in higher education has introduced new opportunities for scalable academic support, but existing systems remain largely unaligned with instructional context and pedagogical intent. Current AI tools operate as general-purpose assistants, lacking awareness of course-specific materials, instructor-defined learning objectives, and continuity across student interactions. This limitation contributes to fragmented learning experiences and weak alignment between academic work and career outcomes. This paper introduces Instructor-Grounded Agentic AI, a framework for designing educational AI systems that integrate course-specific knowledge, pedagogical strategies, and longitudinal student interaction. The proposed approach models learning as a context-conditioned and continuous process, combining retrieval-based grounding, structured pedagogical guidance, and agentic reasoning capabilities. A dual-layer architecture is presented, consisting of a course-grounded assistant that extends instructor presence and a longitudinal learning companion that accumulates student work across courses to support career alignment. We formalize the learning interaction as a function of query, instructional context, pedagogical policy, and memory, and define an objective that captures learning depth, continuity, and career relevance. The paper outlines a mixed-methods research design and presents preliminary insights from an initial deployment, demonstrating the feasibility of the approach. The proposed framework advances a shift from general-purpose AI assistance toward instructionally aligned, continuous learning systems, with implications for educational design and workforce readiness.