Computer science as a discipline encompasses the study and creation of computational techniques for problem-solving. Unfortunately, much of the benefit of these techniques to engineering students is lost in translation, in part due to the disconnect between introductory computing courses and the disciplinary courses that introduce problems that could benefit from computational techniques. A considerable body of literature has documented the ineffectiveness of traditional programming courses in transferring usable knowledge to students, specifically where problem-solving skills and computational thinking are concerned. This paper describes a transformative change to the instruction of introductory computing to engineering students, in the context of a single non-major course. In the proposed architecture, a core "lecture" is supplemented with plug-in domain-specific modules and problems that are presented in domain-specific recitation/laboratory sections. The focus of the lecture component of the course is to introduce the core concepts and skills required for computational thinking, including algorithms and software engineering techniques. The recitation/laboratory component encourages and guides the application of these skills to introductory problems in engineering. The objective is to shift the emphasis from the minutia of the syntax of textual programming languages to computational thinking.