College courses with a history of large enrollment sizes, such as General Chemistry, often rely on online homework systems to provide students with practice in applying new concepts to solve problems. Online homework systems offer many potential advantages, including instant feedback to students, adaptive learning capability, and valuable data to instructors that help identify learning obstacles on-the-fly. However, there does not currently exist network infrastructure that allows a global community of online learners to leverage this wealth of data, which may be generated from different online systems, in order to facilitate synchronous interactions, enable higher cognitive skills to be exercised, and enhance team learning in cyberspace. We have recently developed a framework for the creation of a new networking paradigm to build effective online learning communities: Academic Social Networks (ASNs). The framework integrates several key components: problem template engines (PTEs) that generate questions or exercises that test specific learning objectives, a critical skills network (CSN) that established an underlying fingerprint for each problem that is generated, and a virtual classroom environment (VCE) that allows synchronous interactions to take place in order to enable problem solving and team learning in cyberspace. These components act together to create an environment where students can work problems in order to assess mastery of specific learning objectives. Mastery is tracked at various levels of difficulty that are determined by the set of required critical skills needed to solve each problem. In this way, the CSN provides the foundation for which problems can be connected to one another, mastery of learning objectives can be tracked, and specific learning pathways can be analyzed. A student struggling with a problem that is testing a specific learning objective can reach out to the ASN to connect with other students that have demonstrated mastery of that learning objective at the same difficulty level or higher, and that have a track record at effective peer-mentoring, in order to get help. Ultimately, this framework allows for the development of a tool that leverages the power of large enrollments to facilitate on-demand peer mentoring and delivery of custom instruction at scale. This work represents a significant advance in the development of novel online instructional technology that has promise to create new types of effective online learning communities that improve the quality of education. This may have a profound impact on how we connect with students enrolled in the growing massive open online courses (MOOCs) or those enrolled in large gateway courses at a university.