{"id":182,"date":"2026-08-13T17:23:01","date_gmt":"2026-08-13T17:23:01","guid":{"rendered":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/?post_type=chapter&#038;p=182"},"modified":"2026-09-21T15:36:23","modified_gmt":"2026-09-21T15:36:23","slug":"agent-creation-in-course-design","status":"publish","type":"chapter","link":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/chapter\/agent-creation-in-course-design\/","title":{"rendered":"Practical Teaching Applications &amp; Strategies: Agent Creation in Course Design"},"content":{"raw":"Agent\u00a0creation\u00a0in the context\u00a0of\u00a0AI\u00a0as\u00a0a\u00a0course\u00a0design\u00a0partner\u00a0represents\u00a0a shift\u00a0from generic,\u00a0conversational\u00a0AI\u00a0interactions to an\u00a0ecosystem\u00a0of\u00a0specialized, context-aware\u00a0digital\u00a0colleagues\u00a0tailored\u00a0to specific\u00a0instructional\u00a0needs.\u00a0\u00a0The creation of specialized AI agents that support recurring instructional design activities\u00a0that\u00a0are configured with specific goals, instructions, knowledge resources, and task boundaries.\u00a0 It\u00a0enables\u00a0them to function\u00a0as persistent collaborators throughout the course development\u00a0process.\u00a0 These agents can\u00a0assist in\u00a0instructional\u00a0activities such as\u00a0development, learning outcome\u00a0alignment, assessment construction, rubric generation, assignment redesign, accessibility review, content\u00a0organization, and learning analytics interpretation.\u00a0 AI agents allow instructors to devote greater attention to\u00a0higher order\u00a0responsibilities\u00a0such as pedagogical decision-making, fostering student engagement, mentoring learners, and ensuring academic rigor.\u00a0 The development of\u00a0agentic AI\u00a0emphasizes\u00a0agents that can reason, access external tools, maintain memory, and collaborate within structured workflows to achieve complex objectives, making them particularly well suited for\u00a0educational design environments.\r\n\r\nThe AI agent\u00a0within higher education should be viewed as instructional support systems rather than autonomous course designers.\u00a0 Effective agent creation requires\u00a0instructors\u00a0to define clear parameters regarding learning outcomes, disciplinary expectations, ethical\u00a0standards, accessibility requirement\u00a0aligned with accreditation\u00a0stans, and institutional policies.\u00a0 For example, a Syllabus Design Agent may generate draft course structures aligned with accreditation requirements, an Assessment Builder Agent may\u00a0propose quiz banks and grading rubrics, a Student Persona Agent\u00a0may simulate learner perspectives to identify areas of confusion, and an Accessibility Review Agent may evaluate materials against universal design principles.\u00a0 Therefore, the ultimate responsibility for education quality, academic integrity, student success, and curriculum alignment remains\u00a0with the instructor.\u00a0 This\u00a0approach reflects emerging practices in AI-enhanced education, where AI serves as a collaborative partner that augments rather than replaces\u00a0instructor expertise.\u00a0 Educational redesign initiatives have already demonstrated\u00a0the use of AI collaborators for assignment revision, rubric creation, and peer review support, illustrating how specialized agents can help instructors improve efficiency while maintaining pedagogical control.\r\n<h1>A Framework for Creating Educational AI Agents<\/h1>\r\n<ol>\r\n \t<li><strong>Define the Agent\u2019s Purpose<\/strong>, which is identify a specific instructional design problem the agent will solve.<\/li>\r\n<\/ol>\r\n<em>Example:\u00a0<\/em>\r\n<ul>\r\n \t<li><em>Syllabus Design Agent \u2013 Creates draft syllabi\u00a0aligned with accreditation\u00a0standards.\u00a0<\/em><\/li>\r\n \t<li><em>Assessment Builder Agent \u2013 Generates quizzes, test banks,\u00a0rubrics.\u00a0<\/em><\/li>\r\n \t<li><em>Learning Outcome Alignment Agent\u00a0\u2013 Maps activities and\u00a0assessments to Bloom\u2019s Taxonomy.\u00a0<\/em><\/li>\r\n \t<li><em>Accessibility Review\u00a0Agent\u00a0\u2013\u00a0Reviews materials against University Design for Learning (UDL)\u00a0principles.\u00a0<\/em><\/li>\r\n \t<li><em>Student Persona Agent \u2013 Simulates student perspective to identify confusion points.\u00a0<\/em><\/li>\r\n<\/ul>\r\n&nbsp;\r\n<ol start=\"2\">\r\n \t<li><strong>Establish Agent Instructions<\/strong>,\u00a0which\u00a0is\u00a0creating\u00a0a detailed system prompt that defines\u00a0role, objectives, scope, constraints,\u00a0educational philosophy,\u00a0institutional\u00a0standards.<\/li>\r\n<\/ol>\r\n<em>Example:\u00a0\u00a0An instructional design agent, your\u00a0role is to help faculty create learner-centered course materials.\u00a0 The recommendations must\u00a0align with Bloom\u2019s Taxonomy, active-learning principles, accessibility\u00a0guidelines, and course learning outcomes.\u00a0 It may suggest improvements but never make final pedagogical decisions.\u00a0<\/em>\r\n\r\n&nbsp;\r\n<ol start=\"3\">\r\n \t<li><strong>Providing Knowledge Resources<\/strong>,\u00a0the AI agent should emphasize\u00a0embedding pedagogical knowledge and learning science principles directly into the agent\u2019s\u00a0architecture rather than relying on the user prompting.<\/li>\r\n<\/ol>\r\nThe\u00a0agent should be equipped with following objectives:\r\n<ul>\r\n \t<li>Course\u00a0Learning Outcome<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Program\u00a0Outcomes<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Accreditation\u00a0Requirements<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Institutional Policies<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Accessibility Guidelines<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Assessment Standards<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Discipline-specific Resources<\/li>\r\n<\/ul>\r\n&nbsp;\r\n<ol start=\"4\">\r\n \t<li><strong>Define Task Boundaries<\/strong>, the creation of AI agent should specify what it\u00a0can and cannot do.<\/li>\r\n<\/ol>\r\n<table>\r\n<tbody>\r\n<tr>\r\n<td><strong>Can Do<\/strong><\/td>\r\n<td><strong>Cannot Do<\/strong><\/td>\r\n<\/tr>\r\n<tr>\r\n<td>\r\n<ul>\r\n \t<li>Generate drafts<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Suggest activities<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Recommend assessment strategies<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Analyze curriculum alignment<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Identify content gaps<\/li>\r\n<\/ul>\r\n<\/td>\r\n<td>\r\n<ul>\r\n \t<li>Authorizing\u00a0curriculum\u00a0changes<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Replace faculty judgment<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Assign grades autonomously<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Making\u00a0policy decisions<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Override institutional requirements<\/li>\r\n<\/ul>\r\n<\/td>\r\n<\/tr>\r\n<\/tbody>\r\n<\/table>\r\nThis chart categorizes the boundaries of what tasks AI agents can and cannot perform.\r\n\r\nHuman oversight\u00a0remains critical because AI serves as a collaborator rather than a replacement for instructional expertise.\r\n\r\n&nbsp;\r\n<ol start=\"5\">\r\n \t<li><strong>Design a Human-in-the-loop Workflow<\/strong>,\u00a0the AI agent\u00a0augment cognitive workload\u00a0while educators focus on strategic, creative, and ethical decisions.<\/li>\r\n<\/ol>\r\nThe workflow\u00a0sequence:\r\n<ol>\r\n \t<li>Faculty goal<\/li>\r\n \t<li>AI agent draft<\/li>\r\n \t<li>\u00a0Faculty Review<\/li>\r\n \t<li>Revision &amp; Feedback<\/li>\r\n \t<li>Final Approval<\/li>\r\n \t<li>Course Implementation.<\/li>\r\n<\/ol>\r\n&nbsp;\r\n<ol start=\"6\">\r\n \t<li><strong>Evaluate and Improve the Agent<\/strong>, which refine instructions\u00a0and knowledge sources based on result.<\/li>\r\n<\/ol>\r\nThe AI agent\u00a0should measure the following\u00a0objectives:\r\n<ul>\r\n \t<li>Accuracy<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Pedagogical alignment<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Accessibility compliance<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Bias reduction<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Faculty satisfaction<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Time savings<\/li>\r\n<\/ul>\r\n<h1>Assessment Builder Agent\u00a0Example<\/h1>\r\nGoal: Create assessments aligned with course outcomes.\r\n\r\nInputs:\r\n<ul>\r\n \t<li>Learning outcomes<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Course content<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Bloom\u2019s taxonomy<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Assessment type<\/li>\r\n<\/ul>\r\nOutputs:\r\n<ul>\r\n \t<li>Quiz questions<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Rubric<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Assignment descriptions<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Feedback prompts<\/li>\r\n<\/ul>\r\nHuman Review\r\n<ul>\r\n \t<li>Verify accuracy<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Ensure rigor<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Check academic integrity<\/li>\r\n<\/ul>\r\n<ul>\r\n \t<li>Confirm\u00a0fairness and inclusivity<\/li>\r\n<\/ul>","rendered":"<p>Agent\u00a0creation\u00a0in the context\u00a0of\u00a0AI\u00a0as\u00a0a\u00a0course\u00a0design\u00a0partner\u00a0represents\u00a0a shift\u00a0from generic,\u00a0conversational\u00a0AI\u00a0interactions to an\u00a0ecosystem\u00a0of\u00a0specialized, context-aware\u00a0digital\u00a0colleagues\u00a0tailored\u00a0to specific\u00a0instructional\u00a0needs.\u00a0\u00a0The creation of specialized AI agents that support recurring instructional design activities\u00a0that\u00a0are configured with specific goals, instructions, knowledge resources, and task boundaries.\u00a0 It\u00a0enables\u00a0them to function\u00a0as persistent collaborators throughout the course development\u00a0process.\u00a0 These agents can\u00a0assist in\u00a0instructional\u00a0activities such as\u00a0development, learning outcome\u00a0alignment, assessment construction, rubric generation, assignment redesign, accessibility review, content\u00a0organization, and learning analytics interpretation.\u00a0 AI agents allow instructors to devote greater attention to\u00a0higher order\u00a0responsibilities\u00a0such as pedagogical decision-making, fostering student engagement, mentoring learners, and ensuring academic rigor.\u00a0 The development of\u00a0agentic AI\u00a0emphasizes\u00a0agents that can reason, access external tools, maintain memory, and collaborate within structured workflows to achieve complex objectives, making them particularly well suited for\u00a0educational design environments.<\/p>\n<p>The AI agent\u00a0within higher education should be viewed as instructional support systems rather than autonomous course designers.\u00a0 Effective agent creation requires\u00a0instructors\u00a0to define clear parameters regarding learning outcomes, disciplinary expectations, ethical\u00a0standards, accessibility requirement\u00a0aligned with accreditation\u00a0stans, and institutional policies.\u00a0 For example, a Syllabus Design Agent may generate draft course structures aligned with accreditation requirements, an Assessment Builder Agent may\u00a0propose quiz banks and grading rubrics, a Student Persona Agent\u00a0may simulate learner perspectives to identify areas of confusion, and an Accessibility Review Agent may evaluate materials against universal design principles.\u00a0 Therefore, the ultimate responsibility for education quality, academic integrity, student success, and curriculum alignment remains\u00a0with the instructor.\u00a0 This\u00a0approach reflects emerging practices in AI-enhanced education, where AI serves as a collaborative partner that augments rather than replaces\u00a0instructor expertise.\u00a0 Educational redesign initiatives have already demonstrated\u00a0the use of AI collaborators for assignment revision, rubric creation, and peer review support, illustrating how specialized agents can help instructors improve efficiency while maintaining pedagogical control.<\/p>\n<h1>A Framework for Creating Educational AI Agents<\/h1>\n<ol>\n<li><strong>Define the Agent\u2019s Purpose<\/strong>, which is identify a specific instructional design problem the agent will solve.<\/li>\n<\/ol>\n<p><em>Example:\u00a0<\/em><\/p>\n<ul>\n<li><em>Syllabus Design Agent \u2013 Creates draft syllabi\u00a0aligned with accreditation\u00a0standards.\u00a0<\/em><\/li>\n<li><em>Assessment Builder Agent \u2013 Generates quizzes, test banks,\u00a0rubrics.\u00a0<\/em><\/li>\n<li><em>Learning Outcome Alignment Agent\u00a0\u2013 Maps activities and\u00a0assessments to Bloom\u2019s Taxonomy.\u00a0<\/em><\/li>\n<li><em>Accessibility Review\u00a0Agent\u00a0\u2013\u00a0Reviews materials against University Design for Learning (UDL)\u00a0principles.\u00a0<\/em><\/li>\n<li><em>Student Persona Agent \u2013 Simulates student perspective to identify confusion points.\u00a0<\/em><\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<ol start=\"2\">\n<li><strong>Establish Agent Instructions<\/strong>,\u00a0which\u00a0is\u00a0creating\u00a0a detailed system prompt that defines\u00a0role, objectives, scope, constraints,\u00a0educational philosophy,\u00a0institutional\u00a0standards.<\/li>\n<\/ol>\n<p><em>Example:\u00a0\u00a0An instructional design agent, your\u00a0role is to help faculty create learner-centered course materials.\u00a0 The recommendations must\u00a0align with Bloom\u2019s Taxonomy, active-learning principles, accessibility\u00a0guidelines, and course learning outcomes.\u00a0 It may suggest improvements but never make final pedagogical decisions.\u00a0<\/em><\/p>\n<p>&nbsp;<\/p>\n<ol start=\"3\">\n<li><strong>Providing Knowledge Resources<\/strong>,\u00a0the AI agent should emphasize\u00a0embedding pedagogical knowledge and learning science principles directly into the agent\u2019s\u00a0architecture rather than relying on the user prompting.<\/li>\n<\/ol>\n<p>The\u00a0agent should be equipped with following objectives:<\/p>\n<ul>\n<li>Course\u00a0Learning Outcome<\/li>\n<\/ul>\n<ul>\n<li>Program\u00a0Outcomes<\/li>\n<\/ul>\n<ul>\n<li>Accreditation\u00a0Requirements<\/li>\n<\/ul>\n<ul>\n<li>Institutional Policies<\/li>\n<\/ul>\n<ul>\n<li>Accessibility Guidelines<\/li>\n<\/ul>\n<ul>\n<li>Assessment Standards<\/li>\n<\/ul>\n<ul>\n<li>Discipline-specific Resources<\/li>\n<\/ul>\n<p>&nbsp;<\/p>\n<ol start=\"4\">\n<li><strong>Define Task Boundaries<\/strong>, the creation of AI agent should specify what it\u00a0can and cannot do.<\/li>\n<\/ol>\n<table>\n<tbody>\n<tr>\n<td><strong>Can Do<\/strong><\/td>\n<td><strong>Cannot Do<\/strong><\/td>\n<\/tr>\n<tr>\n<td>\n<ul>\n<li>Generate drafts<\/li>\n<\/ul>\n<ul>\n<li>Suggest activities<\/li>\n<\/ul>\n<ul>\n<li>Recommend assessment strategies<\/li>\n<\/ul>\n<ul>\n<li>Analyze curriculum alignment<\/li>\n<\/ul>\n<ul>\n<li>Identify content gaps<\/li>\n<\/ul>\n<\/td>\n<td>\n<ul>\n<li>Authorizing\u00a0curriculum\u00a0changes<\/li>\n<\/ul>\n<ul>\n<li>Replace faculty judgment<\/li>\n<\/ul>\n<ul>\n<li>Assign grades autonomously<\/li>\n<\/ul>\n<ul>\n<li>Making\u00a0policy decisions<\/li>\n<\/ul>\n<ul>\n<li>Override institutional requirements<\/li>\n<\/ul>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>This chart categorizes the boundaries of what tasks AI agents can and cannot perform.<\/p>\n<p>Human oversight\u00a0remains critical because AI serves as a collaborator rather than a replacement for instructional expertise.<\/p>\n<p>&nbsp;<\/p>\n<ol start=\"5\">\n<li><strong>Design a Human-in-the-loop Workflow<\/strong>,\u00a0the AI agent\u00a0augment cognitive workload\u00a0while educators focus on strategic, creative, and ethical decisions.<\/li>\n<\/ol>\n<p>The workflow\u00a0sequence:<\/p>\n<ol>\n<li>Faculty goal<\/li>\n<li>AI agent draft<\/li>\n<li>\u00a0Faculty Review<\/li>\n<li>Revision &amp; Feedback<\/li>\n<li>Final Approval<\/li>\n<li>Course Implementation.<\/li>\n<\/ol>\n<p>&nbsp;<\/p>\n<ol start=\"6\">\n<li><strong>Evaluate and Improve the Agent<\/strong>, which refine instructions\u00a0and knowledge sources based on result.<\/li>\n<\/ol>\n<p>The AI agent\u00a0should measure the following\u00a0objectives:<\/p>\n<ul>\n<li>Accuracy<\/li>\n<\/ul>\n<ul>\n<li>Pedagogical alignment<\/li>\n<\/ul>\n<ul>\n<li>Accessibility compliance<\/li>\n<\/ul>\n<ul>\n<li>Bias reduction<\/li>\n<\/ul>\n<ul>\n<li>Faculty satisfaction<\/li>\n<\/ul>\n<ul>\n<li>Time savings<\/li>\n<\/ul>\n<h1>Assessment Builder Agent\u00a0Example<\/h1>\n<p>Goal: Create assessments aligned with course outcomes.<\/p>\n<p>Inputs:<\/p>\n<ul>\n<li>Learning outcomes<\/li>\n<\/ul>\n<ul>\n<li>Course content<\/li>\n<\/ul>\n<ul>\n<li>Bloom\u2019s taxonomy<\/li>\n<\/ul>\n<ul>\n<li>Assessment type<\/li>\n<\/ul>\n<p>Outputs:<\/p>\n<ul>\n<li>Quiz questions<\/li>\n<\/ul>\n<ul>\n<li>Rubric<\/li>\n<\/ul>\n<ul>\n<li>Assignment descriptions<\/li>\n<\/ul>\n<ul>\n<li>Feedback prompts<\/li>\n<\/ul>\n<p>Human Review<\/p>\n<ul>\n<li>Verify accuracy<\/li>\n<\/ul>\n<ul>\n<li>Ensure rigor<\/li>\n<\/ul>\n<ul>\n<li>Check academic integrity<\/li>\n<\/ul>\n<ul>\n<li>Confirm\u00a0fairness and inclusivity<\/li>\n<\/ul>\n","protected":false},"author":557,"menu_order":3,"template":"","meta":{"pb_show_title":"on","pb_short_title":"","pb_subtitle":"","pb_authors":["wcheng"],"pb_section_license":""},"chapter-type":[],"contributor":[82],"license":[],"class_list":["post-182","chapter","type-chapter","status-publish","hentry","contributor-wcheng"],"part":110,"_links":{"self":[{"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/pressbooks\/v2\/chapters\/182","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/pressbooks\/v2\/chapters"}],"about":[{"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/wp\/v2\/types\/chapter"}],"author":[{"embeddable":true,"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/wp\/v2\/users\/557"}],"version-history":[{"count":9,"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/pressbooks\/v2\/chapters\/182\/revisions"}],"predecessor-version":[{"id":268,"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/pressbooks\/v2\/chapters\/182\/revisions\/268"}],"part":[{"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/pressbooks\/v2\/parts\/110"}],"metadata":[{"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/pressbooks\/v2\/chapters\/182\/metadata\/"}],"wp:attachment":[{"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/wp\/v2\/media?parent=182"}],"wp:term":[{"taxonomy":"chapter-type","embeddable":true,"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/pressbooks\/v2\/chapter-type?post=182"},{"taxonomy":"contributor","embeddable":true,"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/wp\/v2\/contributor?post=182"},{"taxonomy":"license","embeddable":true,"href":"https:\/\/pressbooks.ulib.csuohio.edu\/usingaiinacademics\/wp-json\/wp\/v2\/license?post=182"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}