Choosing a career is a high-stakes decision, yet students in rural areas often make it with little structured information, limited access to trained counsellors, and few examples of the professions open to them. Computer-based decision support and recommender techniques have been proposed as a way to widen access to guidance, but the literature on them is scattered across psychology, information systems, and machine learning. This paper reviews decision support and recommendation approaches used in career guidance and examines how well they suit the conditions of rural learners in India. Foundational career-development theories, rule-based and multi-criteria decision support, content-based, collaborative, and hybrid recommendation, and supervised machine learning models are compared on their data needs, transparency, and robustness when data are sparse, with particular attention to a recent systematic review of career recommendation systems and a machine-learning study of student career prediction. The review indicates that most existing systems assume rich user histories, reliable connectivity, and English-language interfaces, assumptions that rarely hold in rural schools. Hybrid designs that combine interest- and aptitude-based profiling with transparent rules and multi-criteria ranking appear most defensible where data are scarce. Building on this synthesis, we outline a layered decision support framework for improving career awareness among rural students, covering profiling, knowledge, recommendation, explanation, and counsellor-in-the-loop components. The paper closes with open issues in validation, fairness, and localisation that future empirical work should address.