Applied Artificial Intelligence
Applied AI is no longer a survey of separate perception techniques: practitioners now assemble vision, speech, and language capability out of pretrained and hosted models, wire them together with tool-calling agents and retrieval, and are judged on evaluation, cost, and accountability rather than on hand-built pipelines. The course therefore moves from demonstrating techniques to specifying, orchestrating, evaluating, and taking responsibility for working AI systems.
Current description → proposed description
Applied Artificial Intelligence presents the concepts of intelligence, both human and machine, and the nature of information, its origin, description, and transmission. This course will offer a practical approach to incorporating artificial intelligence into useful applications. It includes such topics as: face recognition, speech recognition, and robotic construction. The nature of human intelligence and the limits of machine intelligence will be treated from a scientific, philosophical, and computational perspective.
This course provides a hands-on introduction to the practical implementation of Artificial Intelligence across diverse domains. Students explore the foundations and limitations of machine intelligence in contrast to human cognition, integrating scientific, computational, and philosophical perspectives. Emphasis is placed on the real-world application of AI models and techniques, including computer vision, natural language processing, speech recognition, and robotics. Through project-based learning, students design and deploy AI solutions using modern frameworks. Key topics include neural networks, large language models, deep learning, reinforcement learning, and edge-AI integration. Students also engage in critical discussions about the boundaries of machine learning, the nature of intelligence, and the ethical responsibilities involved in building autonomous systems. By the end of the course, students will have developed working prototypes that demonstrate how AI can be used to solve complex, real-world problems. Because production AI is now routinely assembled and operated with coding and tool-calling agents, students also specify, orchestrate, and evaluate agentic workflows—retrieval-augmented generation, tool use, guardrails, and evaluation harnesses—and remain accountable for the behavior of the systems those agents help them build.
What changes
- Replaces the short legacy catalog text with the department's fuller workbook description
- Adds neural networks, large language models, deep learning, reinforcement learning and edge AI
- Shifts from topic survey to project-based prototypes built on modern frameworks
- Adds agentic practice: tool calling, retrieval augmentation, guardrails and evaluation harnesses
- Names ethical responsibility for autonomous systems as graded course content
6 proposed outcomes, mapped to 12 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 differentiate machine intelligence from human cognition by testing semantic, historical, and popular definitions of artificial intelligence against the measured behavior, capability ceilings, and characteristic failure modes of contemporary learning and agentic systems.
Testing semantic, historical, and popular definitions of artificial intelligence against the measured behavior and characteristic failure modes of contemporary systems is exactly the defining-and-evaluating work PLO 5.2 names.
Students will be able to implement applied AI solutions for computer vision, speech recognition, and natural language tasks by training and adapting supervised, unsupervised, and reinforcement-learning models on data they have themselves collected, cleaned, and documented from APIs, sensors, and stored sources.
Training and adapting supervised, unsupervised, and reinforcement-learning models is the system-building evidence for PLO 4.1 and the direct application of the neural, probabilistic, and search foundations PLO 3.1 names, since training a network, fitting a probabilistic model, or learning a policy is that theory applied; the requirement that students collect, clean, and document their own data from APIs, sensors, and stored sources is the data-acquisition evidence for PLO 2.1.
Students will be able to construct a domain-specific AI agent that calls external tools and retrieves from a curated knowledge source at run time, engineering the prompts, output schemas, and guardrails that keep its task execution scoped, secure, and resistant to prompt injection.
Building an agent with run-time tool calling and retrieval over a curated source is the secure, real-time-knowledge integration PLO 2.2 describes, and the prompts, output schemas, and injection guardrails the student must author are the prompt-engineering evidence for PLO 4.2.
Students will be able to evaluate a working AI prototype against task-appropriate metrics and an agent evaluation harness—held-out benchmarks, LLM-as-judge scoring, and stated cost and latency budgets—using the diagnostic results to direct successive rounds of model, prompt, and retrieval refinement.
Measuring a prototype with task-appropriate metrics and then driving successive refinement from the diagnostics is precisely the assess-and-iteratively-refine cycle of PLO 6.2, and the prompt revisions that harness forces are graded prompt-engineering work under PLO 4.2.
Students will be able to integrate a large language model into an enterprise or edge deployment target through hosted APIs, embeddings, and vector retrieval under explicit latency, cost, and memory budgets, documenting the resulting architecture, its cost and latency trade-offs, and its ethical and accountability implications for both engineering and non-technical stakeholders.
Deploying an LLM to an enterprise or edge target via APIs, embeddings, and vector retrieval under real budgets is the industrial and embedded deployment PLO 6.1 names, and the required dual-audience document must carry the ethical and accountability implications as well as the design and its results, which is the full concepts-results-and-ethics evidence PLO 5.1 specifies.
Students will be able to critique the ethical, legal, and vocational consequences of an autonomous system they have built—bias, transparency, accountability, data and model provenance, and human-in-the-loop escalation—defending the limits they place on its autonomy as service to the people it affects.
Critiquing bias, transparency, accountability, and provenance in a system the student personally built is the ethical-and-legal-integrity evidence for the outcome the workbook states twice in identical words, once under Christian Faith as PLO 1.1 and once under Integrated Disciplinary Knowledge as PLO 3.2; defending the autonomy limits as service to the people affected is the vocational articulation required by PLO 1.2.
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 twelve.
The workbook maps this course too
The proposal workbook's assessment map already assigns program learning outcomes to this course. The outcomes above were written to cover it.
| Program outcome | Workbook level | In this draft |
|---|---|---|
| PLO 1.1 Ethical and Legal Integrity in AI | I | covered |
| PLO 1.2 Vocation and Christian Worldview | I | covered |
| PLO 2.1 Data Acquisition for Real Problems | I | covered |
| PLO 2.2 Domain-Specific AI Agents | I | covered |
| PLO 3.1 AI Theory and Mathematics | D,AE | covered |
| PLO 3.2 Ethical and Legal Integrity in AI | I | covered |
| PLO 4.1 Learning-Based Systems and Agents | D,AE | covered |
| PLO 4.2 Prompt Engineering for LLMs | D,AE | covered |
| PLO 5.1 Communicating AI to Any Audience | D,AE | covered |
| PLO 5.2 Defining and Evaluating AI | D,AE | covered |
| PLO 6.1 LLMs in Industrial Settings | D | covered |
| PLO 6.2 Model Evaluation and Refinement | D,AE | covered |
I = Introduced · D = Developed · AE = Assessed at Exit