There is a growing interest in developing students' interest in computer science and computational thinking as computing has become ubiquitous in various fields. This has led to a plethora of online educational platforms offering computer science content for learning. There are several advantages and disadvantages of existing platforms. To address the current limitation of both the paid and free online learning activities, we present the Curated Pathways to Innovation (CPI), an innovative platform that utilizes machine learning techniques to curate and personalize readily available computational learning material online. Initial results suggest that students benefit from using CPI. Not only are students exposed to developmentally appropriate and personally salient computer science-related content aligned with educational standards, but they also experience a significant increase in their attitudes towards computer science activities, particularly those involving computer programming and computational thinking. In this way, CPI is highly scalable, having the potential to reach a broad audience of learners by curating content from an integrated set of educational resources and thus also orienting a future generation of students towards careers in computer science and related fields. The strengths of this approach, as well as opportunities for future platform, content, and curriculum development, are considered.