Software Engineering
Software engineering is the discipline most directly reshaped by agentic AI: a large share of production code is now drafted by coding agents, moving the professional's work toward specification, decomposition, review of machine-generated output, verification, and accountability for code the engineer did not type. The course keeps its lifecycle-management identity but retrains its core practices - requirements, design review, quality assurance, and team process - around directing and verifying agent-assisted work.
Current catalog prerequisites — (CSC 250 or 2050).
Current description → proposed description
This course is the management of the entire software development process. This course affords the student the opportunity to explore the art and science of professional software development in great detail. The foundational aspects of the creative process, idea, implementation, and interaction are investigated in the context of software development. Principles of requirements, specifications, design, implementation, and maintenance are studied. The software development lifecycle is used as a management tool for the professional creation of effective systems. Support and management issues including design patterns, user and developer documentation, coding tools, and quality assurance are investigated. Actual programming projects are analyzed along with current research in the field. Several major software projects, both individual and team, are synthesized by students using an industry methodology. Knowledge of the programming environment utilized in CSC 2050 is required. CSC 3070 is part of the AI concentration in the CS curriculum. CSC 3070 may satisfy university requirements as a Writing Intensive course.
Prerequisites: (CSC 250 or 2050).
This course examines the management of the entire software development process at a time when much production code is drafted, reviewed, and operated with the assistance of AI coding agents. Students study requirements elicitation, specification, architecture and design, implementation, verification, and maintenance across the software development lifecycle, and carry individual and team projects through an industry methodology. Emphasis is placed on the practices that now distinguish the professional engineer: writing specifications precise enough to direct an agent, decomposing work into reviewable units, critically reviewing machine-generated code and tests, and retaining accountability for software the engineer did not personally type. Design patterns, version control and code review workflows, continuous integration, evaluation and regression harnesses, user and developer documentation, and quality assurance are investigated alongside agent-specific concerns including provenance and licensing, prompt injection and supply-chain risk in development toolchains, and cost, latency, and observability budgets. CSC 3070 is part of the AI concentration in the CS curriculum. CSC 3070 may satisfy university requirements as a Writing Intensive course.
What changes
- Specifications and work decomposition written to direct AI coding agents
- Code review reframed as verification of machine-generated code and tests
- Evaluation harnesses and continuous integration added to quality assurance
- Provenance, licensing, prompt injection, and toolchain supply-chain risk
- Explicit engineer accountability for software the engineer did not type
6 proposed outcomes, mapped to 10 program outcomes
Each outcome below is written to be observable and assessable, and each is mapped to the program learning outcomes for which it produces evidence.
Students will be able to construct written requirements specifications, acceptance criteria, and work decompositions precise enough to direct both human developers and AI coding agents.
PLO 5.1: the graded artifact is written specification and acceptance criteria that must be unambiguous enough to direct an agent, which is precisely the transparent, integrity-bearing written communication PLO 5.1 names.
Students will be able to design a software architecture that applies established design patterns, documenting the trade-offs among maintainability, performance, and operating cost when portions of the system are generated or operated by AI components.
PLO 4.2: the CLO requires proposing an architecture in which AI components carry part of the system's work and defending that choice against maintainability, performance, and cost trade-offs, matching 'propose and evaluate AI-driven solutions to complex issues.'
Students will be able to evaluate AI-generated code, tests, and documentation against specification, security, and licensing criteria to reach a documented accept, revise, or reject decision.
PLO 4.1: critically judging machine-produced output against specification and security criteria, and defending a documented accept/revise/reject decision, is the critical analysis of data-driven outcomes and their implications that 4.1 describes.
Students will be able to implement a quality-assurance pipeline that combines unit, integration, and regression testing with continuous integration and evaluation harnesses to measure the reliability of agent-assisted contributions.
PLO 6.1: the pipeline exists to produce and explain reliability results for AI-assisted output, which is interpreting and explaining AI data outcomes. PLO 6.2: 'measure the reliability' requires drawing defensible conclusions from regression and evaluation-harness results rather than reporting raw pass counts.
Students will be able to apply an industry software development methodology across a team project, maintaining traceable version control, code review, and provenance records, and conducting iteration retrospectives.
PLO 3.3: the team project balances technical delivery against provenance and accountability obligations, which is the technical/ethical balance 3.3 asks teams to hold, and matches the workbook's own use of 3.3 for the CSC 4900 team project. PLO 4.3: 'conducting iteration retrospectives' is the structured self-reflection used to continuously assess and improve the team's approach.
Students will be able to articulate, in writing and in oral presentation to non-specialist stakeholders, the professional and ethical obligations an engineer retains for AI-assisted software, including disclosure of provenance, risk to users, and the engineer's calling to steward systems that others depend on.
PLO 1.2: the graded work is reflective argument about the obligations an engineer retains, naming disclosure of provenance and risk to users, which is the reflection on how Christian values guide decisions about transparency and accountability that 1.2 requires. PLO 1.3: 'the engineer's calling to steward systems that others depend on' is directly the vocation-and-stewardship claim 1.3 makes about work in AI serving humanity. PLO 5.3: the audience is specified as non-specialist stakeholders, which is the translation of technical knowledge for non-specialists that 5.3 requires.
Program outcomes this course reaches
Filled cells are program learning outcomes with at least one supporting course learning outcome in this course. Sparse coverage is expected — no single course carries all eighteen.