Prescriptive Analytics Systems (PAS) represent the most mature iteration of business analytics, significantly enhancing organizational decision-making. Recently, research has gained traction, with various technological innovations, including machine learning and artificial intelligence, significantly influencing the design of PAS. Although recent studies highlight these developments, the rising trend focuses on broader implications, such as the synergies and delegation between systems and users in organizational decision-making environments. Against this backdrop, we utilized a systematic literature review of 262 articles to build on this evolving perspective. Guided by general systems theory and socio-technical thinking, the concept of an information systems artifact directed this review. Our first objective was to clarify the essential subsystems, identifying 23 constituent components of PAS. Subsequently, we delved into the meta-level design of PAS, emphasizing the synergy and delegation between the human decision-maker and prescriptive analytics in supporting organizational decisions. From this exploration, four distinct system archetypes emerged: advisory, executive, adaptive, and self-governing PAS. Lastly, we engaged with affordance theory, illuminating the action potential of PAS. Our study advances the perspective on PAS, specifically from a broader socio-technical and information systems viewpoint, highlighting six distinct research directions, acting as a launchpad for future research in the domain.

规范性分析系统 (PAS) 代表了最成熟的业务分析迭代，可显着增强组织决策。最近，研究取得了进展，包括机器学习和人工智能在内的各种技术创新对 PAS 的设计产生了重大影响。尽管最近的研究强调了这些发展，但上升的趋势侧重于更广泛的影响，例如组织决策环境中系统和用户之间的协同和授权。在此背景下，我们对 262 篇文章进行了系统的文献综述，以建立这一不断发展的观点。在一般系统理论和社会技术思维的指导下，信息系统工件的概念指导了本次审查。我们的首要目标是澄清基本子系统，确定 PAS 的 23 个组成部分。随后，我们深入研究了 PAS 的元级设计，强调人类决策者和规范分析之间在支持组织决策方面的协同和授权。从这次探索中，出现了四种不同的系统原型：咨询型、执行型、适应性型和自治型 PAS。最后，我们研究了可供性理论，阐明了 PAS 的行动潜力。我们的研究推进了 PAS 的视角，特别是从更广泛的社会技术和信息系统的角度来看，突出了六个不同的研究方向，作为该领域未来研究的启动板。