How would I design a personalized, AI-driven educational system that delivers general education (GE) requirements based on demonstrable skills rather than course credits? This is the question of the hour, as higher ed faces up to the reality that student embrace of AI has fundamentally altered the contract between university and student. It is later than you thought.
The first step is to deconstruct the existing GE curriculum (I am using the California State University system as an example) into a core competency framework, translating such subject areas as “Quantitative Reasoning” and “Arts and Humanities” into specific, measurable skills that students are required to master. The second, more complex step is to build the technological platform that can guide each student along a path to mastering these competencies, driven by personalized interests and goals.
The goal is rather than having students sit through an expensive sequence of state-mandated classes, each with a state mandated “learning outcome,” they would engage with AI, in coordination with faculty mentors, to learn what the state demands they learn. And rather than having AI do the work (and not learning anything) they would work with AI to actually learn.
I understand this is bold. But the current system is not working.
I am asked regularly: “how.” The technical foundation for this system would be a microservices architecture. Instead of building a single, massive application, this approach structures the platform as a collection of smaller, independent services that communicate with each other. Think of it like building with LEGOs. Each service handles one specific job — like managing student profiles or recommending projects — and can be developed, updated, or scaled on its own without affecting the entire system. The idea is to make a resilient platform, where a failure in one service would not bring down the others, which would be agile, allowing for rapid integration of new technologies and pedagogical tools.
My proposed CSU GE-delivery architecture would be composed of several key services working in concert, to scale up for a million learners. A Learner Profile Service would create and maintain a dynamic profile for each student, capturing both their academic progress and their interests, goals, and learning preferences. The LPS would engage in direct conversation and also track student engagement. This profile would feed into a Project Recommendation Engine, which would generate personalized, interdisciplinary projects (or “Quests”) that align with the student’s interests while targeting the specific competencies they need to develop.
A Conversational AI would act as the primary interface, a 24/7 tutor and guide that uses Socratic dialogue to help students, answer questions, and provide just-in-time learning resources curated by a dedicated Content Delivery Service.
As students complete these projects, they would submit their work, which could be anything from a research paper to a software prototype or a video documentary, to a digital ePortfolio. An Assessment and Feedback Service would then use AI for initial formative feedback, while also managing a workflow for peer review and final validation by human mentors (faculty) against clear, competency-based rubrics. This creates a continuous loop of learning, reflection, and demonstration.
In short, the GE delivery system will be a combination of AI and faculty. The AI will handle the delivery of information and initial feedback. Faculty time is freed up to act as high-level mentors and coaches who engage with students on deeper conceptual challenges.
Does this seem huge and complicated? No more than the current structure of thousands of human teachers of widely varying qualifications, thousands of classrooms, hundreds of millions of dollars, an alarming non-completion rate, and an epic lack of quality control.
Step 1: Establish the Competency Framework
This initial step involves translating the existing California State University (CSU) General Education (GE) requirements into a functional, skill-based framework.
Analyze Current GE Structure: First, my proposed system will review the CSU GE-Breadth requirements (Areas A-F) and the new Cal-GETC transfer curriculum. Identify all mandated subject areas, subareas, and specific requirements, such as the “Golden Four” (Oral Communication, Written Communication, Critical Thinking, and Quantitative Reasoning).
Extract Core Learning Outcomes: The system will synthesize the official Student Learning Outcomes (SLOs) published by various CSU campuses for each GE area. The goal is to move beyond course titles and identify the specific, demonstrable skills students are expected to acquire (e.g., “construct and deliver persuasive oral messages,” “apply the scientific method,” “analyze and interpret creative expression”).
Develop a Competency Framework: Reorganize the extracted SLOs into six integrated Core Competency Domains. This creates a new operational lexicon focused on ability rather than course completion. The domains are:
Communication & Argumentation (Area A)
Quantitative & Scientific Reasoning (Area B)
Artistic & Humanistic Inquiry (Area C)
Social & Institutional Analysis (Area D & US Government)
Lifelong & Personal Development (Area E)
Cultural & Equity Fluency (Area F)
Create a Competency Map: Construct a detailed mapping table that serves as a "Rosetta Stone," explicitly linking each traditional CSU GE subarea to its corresponding SLO, the relevant Core Competency Domain, a specific sub-competency, and examples of how a student might demonstrate mastery.
Table 1: The Competency Framework Map (selection here; balance in footnotes)1
Step 2: Design the System Architecture
This step outlines the technical blueprint for the AI-powered platform, based on the principles of Intelligent Tutoring Systems (ITS) and a microservices architecture.
Adopt an ITS Model: Structure the system around the four classic components of an ITS:
Domain Model: The AI Competency Framework (from Step 1).
Student Model: A Dynamic Learner Profile for each student.
Tutoring Model: An engine for generating projects and providing feedback.
User Interface: A conversational AI and ePortfolio dashboard.
Implement a Microservices Architecture: Build the platform as a collection of small, independent, and scalable services that communicate via APIs. This approach ensures resilience, agility for updates, and allows for the use of the best technology for each function.
Develop Core Microservices: Create the specific services required for platform functionality.
Table 2: Microservices Architecture
Step 3: Implement the Onboarding Process and Learner Profile
This step focuses on creating the initial student experience and the data model that drives personalization.
Design a Conversational Onboarding: Replace traditional admissions forms with a guided, Socratic dialogue managed by the Conversational AI. Use open-ended prompts to discover a student’s intrinsic motivations, interests, and goals.
Construct the Dynamic Learner Profile: Create a multi-dimensional, evolving profile for each student that captures four key areas:
Interests, Aspirations, and Drivers: Explicitly captured from conversational inputs.
Academic State: Captured via optional transcript uploads and low-stakes knowledge probes.
Learning Preferences: Captured through direct questions and implicit tracking of how students interact with different content types (e.g., video, text, simulations).
Demographics and Context: Voluntary information about a student’s study environment and access to technology to ensure equitable support.
Develop Data Inference Capabilities: Program the system to synthesize data and infer deeper insights.
Topic Modeling: Use NLP techniques to identify latent topics of interest from all student interactions.
Behavioral Analysis: Analyze how students use the platform to infer interests that may not be explicitly stated.
Competency Pattern Recognition: Identify patterns of strength and weakness across competencies to inform future project recommendations.
Step 4: Develop the Project-Based Learning System
This step details the core instructional methodology where students acquire competencies through projects.
Adopt a Project-Based Learning (PBL) Model: Structure the curriculum around CSU-faculty-approved “main course” projects, where the project is the primary vehicle for learning, not a final activity. This approach naturally fosters critical thinking, collaboration, and self-directed learning.
Create the “Quest” System: Use the Project Generation & Recommendation Engine to create and manage personalized projects called “Quests.”
Recommendation: The engine analyzes the student’s Learner Profile to find the intersection of their interests and competency gaps, then recommends relevant Quests.
Co-Design: The student works with the AI to refine the project's driving question, final product format, and milestones, fostering ownership.
Ensure Interdisciplinary Design: Design Quests to inherently require the integration of skills from multiple competency domains, mirroring real-world problem-solving. For example, a Quest to design a sustainable tiny house community would require Quantitative Reasoning (cost/energy calculations), Social Analysis (zoning laws, community needs), and Communication (written proposal, video pitch).
Implement Just-in-Time Scaffolding: Use the Content Curation & Delivery Service to provide small, targeted learning resources ("nuggets") at the moment a student encounters a challenge within a Quest. This microlearning approach provides precisely the information needed to overcome a hurdle and continue the project.
Step 5: Create the Assessment and ePortfolio System
This step establishes the method for demonstrating and validating competency mastery.
Build the Living ePortfolio: Develop an ePortfolio platform that serves as the central archive for all student work. Design it to support three functions:
Developmental Portfolio: A private workspace capturing the entire learning process, including drafts and feedback.
Assessment Portfolio: A curated collection of best work submitted for formal competency validation.
Showcase Portfolio: A public-facing version for sharing with employers or graduate schools.
Support Multi-Modal Artifacts: Design the system to accept and display a wide variety of evidence types, including text documents, data visualizations, videos, audio files, and software prototypes. Require a writtenreflective statement for each artifact, where the student explains how it demonstrates specific competencies.
Develop Competency-Based Rubrics: Create a library of standardized rubrics, one for each sub-competency in the Framework. These rubrics would be:
Aligned to Competencies: The same rubric is used to assess a competency regardless of the project or artifact type.
Developmental: They describe a continuum of performance (e.g., Novice, Developing, Proficient, Mastery) rather than a simple point scale.
Behavior-Focused: They use precise, observable language to describe what a student can do at each level.
Establish a Multi-Layered Feedback Loop: Decouple feedback from final validation by implementing a three-tiered process:
AI Formative Feedback: The system provides immediate, automated feedback on draft artifacts, checking for clarity, structure, and other mechanical aspects.
Structured Peer Feedback: The platform facilitates a guided peer review process where students use the official rubrics to provide feedback to each other.
Mentor Summative Assessment: A human mentor performs the final validation, reviewing the artifact and reflective statement against the rubric to confirm competency achievement.
Step 6: Integrate Human Mentorship and Collaboration
This step ensures the system augments, rather than replaces, human connection and community.
Redefine Educator Roles: Shift human educators from content delivery to higher-value interaction roles:
Mentors: Subject matter experts who provide high-level intellectual guidance and validate final competency mastery.
Coaches: Pedagogical experts who help students with goal-setting, project management, and metacognitive skills.
Community Facilitators: Leaders who organize synchronous group activities, workshops, and discussions to build community.
Deploy the Collaboration & Mentorship Service: Use this dedicated microservice to engineer human connection:
Intelligent Team Formation: Suggests teams for group projects based on complementary skills and interests.
Integrated Tools: Provides built-in video conferencing, whiteboards, and collaborative editing.
Mentor Matching: Connects students with available mentors who have the specific expertise needed for a given problem.
Peer Feedback Hub: Manages the structured peer review workflow.
Foster a Learning Community: Implement features that make learning a visible, social activity:
Quest-Based Forums: Create ad-hoc communities for students working on similar projects to share resources and ask questions.
Project Showcase Galleries: Allow students to publish finished work to public galleries, creating an intellectual commons for inspiration and peer recognition.
Synchronous Events: Host optional, real-time workshops and moderated discussions led by facilitators and mentors.
Step 7: Establish Governance, Ethics, and Accreditation Protocols
This final step addresses the practical implementation challenges to ensure the system is equitable, secure, and viable.
Ensure Equity and Access:
Address the Digital Divide: Pair the system with institutional programs providing device loans and internet access support.
Audit for Algorithmic Bias: Establish an independent AI Ethics Board to regularly audit recommendation and assessment algorithms for fairness.
Promote Accessibility: Use AI tools to provide built-in text-to-speech, translation, and other features to support all learners.
Uphold Academic Integrity:
Design for Authenticity: Focus assessment on complex, personalized, multi-modal projects that are inherently difficult to produce inauthentically with generative AI.
Teach Ethical AI Use: Rather than banning AI, create explicit guidelines for its use as a tool, modeling professional practice.
Validate the Process: Use the developmental ePortfolio to track the learning journey, flagging anomalous submissions (e.g., a final product with no drafts or research) for human review.
Protect Student Data:
Implement Student-Centered Governance: Operate on a principle of explicit, informed consent, giving students control over their data via a clear privacy dashboard.
Ensure Transparency: Adhere to data minimization, collecting only what is essential, and make privacy policies clear and understandable.
Build for Security: Use the microservices architecture for security isolation and encrypt all sensitive data.
Plan for Accreditation:
Provide Superior Evidence: Argue to accreditors that the ePortfolio system provides a more direct and robust body of evidence of student learning than a traditional transcript.
Align with CBE Standards: Design the program to meet the standards of established Competency-Based Education (CBE) certification bodies like Cognia.
Partner with Accreditors: Engage with accrediting agencies from the project's inception, demonstrating transparently how the Competency Framework Map and rubric-based assessments directly fulfill all mandated GE learning outcomes.
Conclusion
It’s a lot. But it’s a lot less than the mess that currently exists. Successful deployment would depend on maintaining consistent standards across the distributed network of mentors and Assessment Services within the microservices architecture. The platform would need to incorporate systematic mentor calibration protocols where faculty achieve validated inter-rater reliability scores before conducting summative assessments against the competency rubrics. The Assessment and Feedback Service should include automated monitoring capabilities that identify significant variations in competency validation patterns, triggering additional calibration interventions when needed. Quality assurance mechanisms such as random portfolio sampling for secondary review and mandatory documentation requirements for borderline assessments will ensure that the personalized learning pathways maintain rigorous academic standards. These operational safeguards preserve the integrity of the Competency Framework while supporting the flexible, student-centered approach that distinguishes this system from traditional course-based models.
The framework outlined here is the beginning of a conversation. The hope would be to provide CSU with a practical roadmap — perhaps it is doing this already, but who knows? Ideally, these conversations would involve forming cross-campus working groups to develop detailed implementation timelines, platform development, technology partners. The first public system to transform GE delivery through a competency-based approach, creating a model that serves students more effectively while maintaining academic rigor and institutional credibility, wins, the way I look at it. Everyone knows it is not working now.
Selected Sources
CSU General Education (GE) | CSU. https://www.calstate.edu/csu-system/administration/academic-and-student-affairs/academic-programs-innovation-and-faculty-development/program-development/Pages/csu-general-education.aspx
CSU – California State University General Education Breadth Requirements:2024-2025 - CONTRA COSTA COLLEGE. https://www.contracosta.edu/wp-content/uploads/2024/07/CSU-GE_2024-2025.pdf
California State University (CSU) General Education (GE) Breadth Certificate of Achievement | Taft College. https://www.taftcollege.edu/academics/programs/csu-general-education-breadth-certificate.php
California State University General Education (GE) Breadth Requirements N IP C - LACC. https://www.lacc.edu/sites/lacc.edu/files/2023-06/LACC%20CSUGE%202023-2024_0.pdf
Guiding Notes for General Education Course Review - Transfer Programs. https://transferprograms.calstate.edu/sites/default/files/2024-01/2022-2023%20GE%20Guiding%20Notes_Final%20September%202022.pdf
General Education (GE) Program Learning Outcomes | California https://www.csub.edu/ge/program-learning-outcomes.shtml
General Education Student Learning Outcomes. https://catalog.csun.edu/general-education/student-learning-outcomes/
General Education - Division of Academic Affairs. https://academics.fresnostate.edu/curriculum/general-education/index.html
General Education Area-Specific Student Learning Outcomes. https://www.csulb.edu/academic-senate/general-education-area-specific-student-learning-outcomes-geslos
GE Learning Outcomes for Students | Sacramento State. https://www.csus.edu/undergraduate-studies/general-education/ge-learning-outcomes-students.html
GE Learning Outcomes for Faculty | Sacramento State/ https://www.csus.edu/undergraduate-studies/general-education/ge-learning-outcomes.html
Competency Frameworks in Higher Education - Akari | Curriculum Management Software. https://akarisoftware.com/2024/04/17/competency-frameworks-in-higher-education/
What Is Competency Mapping? Methods and Why It Matters - Coursera. https://www.coursera.org/enterprise/articles/competency-mapping
General Education and Student Learning Outcomes < Citrus College. https://catalog.citruscollege.edu/programs-study/graduation-requirements-associate-degree/general-education-and-learning-outcomes/
Area E Guidelines (GE 2020) - General Education - Cal Poly, San Luis Obispo. https://ge.calpoly.edu/program/area-e
The General Education Program - California State University, LA. https://ecatalog.calstatela.edu/content.php?catoid=54&navoid=6256
GUIDE TO GE/BREADTH - Cal State East Bay. https://www.csueastbay.edu/ge/breadth-semesters.html
Undergraduate General Education Requirements | California State University, Bakersfield. https://catalog.csub.edu/academic-degrees-programs/undergraduate-general-education-requirements/
Intelligent Tutoring Systems: A Comprehensive Guide. https://www.numberanalytics.com/blog/intelligent-tutoring-systems-ultimate-guide
How a cloud-native and microservices platform can give schools the advantage. https://www.educationhorizons.com/blog/how-a-cloud-native-and-microservices-platform-can-give-small-schools-the-advantage
How Microservices Improve Scalability in Educational Technology. https://colorwhistle.com/microservices-scalability-edtech/
Main components of an Intelligent Tutoring System - ResearchGate. https://www.researchgate.net/figure/Main-components-of-an-Intelligent-Tutoring-System_fig1_321257406
AI Conversation - Blackboard Help. https://help.blackboard.com/Learn/Instructor/Ultra/Interact/AI_Conversation
Connecting Across Disciplines in PBL - Edutopia. https://www.edutopia.org/article/interdisciplinary-project-based-learning/
Personalized Learning: A Teacher's Guide - Education Advanced. https://www.educationadvanced.com/blog/personalized-learning
Learner Profile - Institute for Personalized Learning Application - Digital Promise. https://microcredentials.digitalpromise.org/explore/learner-profile
Inferring user interest | Request PDF - ResearchGate. https://www.researchgate.net/publication/3419437_Inferring_user_interest
How to ensure education equity in the age of AI - Watermark Insights. https://www.watermarkinsights.com/resources/blog/how-to-ensure-education-equity-in-the-age-of-ai/
What is Project Based Learning? | PBLWorks. https://www.pblworks.org/what-is-pbl
The Power of Project-Based Learning (PBL) - CCE Finland. https://www.ccefinland.org/post/project-based-learning-fostering-critical-thinking-and-collaboration-in-education
Project-based learning for STEAM - Pi-Top. https://www.pi-top.com/pbl/for-steam/resources
The Ultimate Personalized Learning Guide - Education Elements. https://www.edelements.com/personalized-learning
Project-Based Learning in STEM: Enhancing Student Engagement. https://onlineprograms.education.uiowa.edu/blog/project-based-learning-in-stem-enhancing-student-engagement
Practical Strategies for Interdisciplinary Teaching in Today's University - Faculty Focus. https://www.facultyfocus.com/articles/effective-teaching-strategies/practical-strategies-for-interdisciplinary-teaching-in-todays-university/
Powering Up: Interdisciplinary Project-Based Learning for Young Engaged Citizens. https://www.naeyc.org/resources/pubs/yc/spring2024/project-based-learning
Project Examples: Teaching STEM Through PBL - SmartLab Learning. https://www.smartlablearning.com/teaching-stem-through-pbl-with-examples/
System Design and Architecture of an Online, Adaptive, and Personalized Learning Platform. http://dspace.mit.edu/handle/1721.1/102982
EPortfolios – Instructional Technology And Design Services - Montclair State University. https://www.montclair.edu/itds/digital-pedagogy/pedagogical-strategies-and-practices/eportfolios/
ePortfolios - AAC&U. https://www.aacu.org/trending-topics/eportfolios
E-Portfolios and Their Uses in Higher Education - Center for Teaching and Learning. https://ctl.risepoint.com/e-portfolios-and-their-uses-in-higher-education/
Using Student Learning Portfolios for Departmental Outcomes Assessment, https://assessment.uiowa.edu/using-student-learning-portfolios-departmental-outcomes-assessment
Using Portfolios in Program Assessment | Assessment and Curriculum Support Center - University of Hawaii at Manoa. https://manoa.hawaii.edu/assessment/resources/using-portfolios-in-program-assessment/
Portfolio Assessment Guidelines - Prior Learning Assessment - Penn State. https://priorlearning.psu.edu/portfolio-assessment-guidelines
Not All Rubrics Are Created Equal: How Competency-Based Progressions Are The Way to Go. https://learnercentered.org/blog/not-all-rubrics-are-created-equal-how-competency-based-progressions-are-the-way-to-go/
Competency-based Rubric Design - Center for Collaborative Education. https://microcredentials.digitalpromise.org/explore/competency-based-rubric-design
General Education Assessment Rubrics | Fairleigh Dickinson University https://www.fdu.edu/academics/general-education-requirements/gen-ed-assessment-rubrics/
Creating Learning Outcomes - Stanford Teaching Commons. https://teachingcommons.stanford.edu/teaching-guides/foundations-course-design/course-planning/creating-learning-outcomes
Guiding Principles for Competency-Based Education and Assessment - American Association of Colleges of Nursing. https://www.aacnnursing.org/Portals/0/PDFs/Essentials/Guiding-Principles-for-CBE-Assessment.pdf
How AI for Education Can Address Digital Equity. https://digitalpromise.org/2024/02/20/how-ai-for-education-can-address-digital-equity/
Provide Equal Access to AI | AAMC. https://www.aamc.org/about-us/mission-areas/medical-education/principles-ai-use/provide-equal-access-ai
Why Equity Must Be a Core Part of the Conversation About AI - Campus Technology. https://campustechnology.com/articles/2024/08/06/why-equity-must-be-a-core-part-of-the-conversation-about-ai.aspx
Strategies and Tools for Academic Integrity in Online Environments | Teaching and Learning Resource Center - The Ohio State University. https://teaching.resources.osu.edu/teaching-topics/strategies-tools-academic-integrity
Analyzing an Artificial Intelligence-Supported Assessment and Student Feedback. https://www.centerforengagedlearning.org/analyzing-an-artificial-intelligence-supported-assessment-and-student-feedback/
All Aboard: The Ethics of Campus AI and Higher Education's New Trolley Problem. https://www.newamerica.org/oti/briefs/new-trolley-problem/
E-Portfolios - Guides - Bonner Wiki. http://bonner.pbworks.com/E-Portfolios+-+Guides
Competency-based Education Certification for Schools - Cognia. https://www.cognia.org/cbe-certification/
I would be interested to see a version of this suited to a private middle and high school.
Possibly easier to introduce at a smaller scale, and starting with students who are just entering academic life.
I am a fan of Hollis Robbins's heterodoxy when it comes to AI and higher education, not because I agree with her, but because Anectotal Value offers ideas that challenge the usual terms of what Henry Farrell calls AI Fight Club, which pits those who see AI-based educational tools as our inevitable future against those who resist AI in large or small ways.
In proposing that CSU automate general education through a microservices architecture consisting of AI applications, Robbins offers a concrete alternative (a genuinely awful one) to the already existing structure of gen-ed (also genuinely awful). Actually existing systematic general education is no longer viable because OpenAI built Shel Silverstein's "Homework Machine" and offers it to college students at no cost. AI, it seems, is the future of the modern state university system when it comes to providing the breadth of knowledge that a liberal arts education is supposed to provide. I appreciate the way the proposal presents a version of that future unflinchingly.
My primary objection is to Step 4: Develop the Project-Based Learning System and Step 6: Integrate Human Mentorship and Collaboration, because they are based on the increasingly dubious notion that credentialing and learning are meaningfully related. We have before us the opportunity to abandon this fiction by automating the former and leaving the latter up to the student to figure out, maybe with a few mentors around to help. Why muddy the waters by attempting to automate what remains of learning?
Far better, I say, to create loose structures of in-person connections where students who are interested pursue their interests via projects under the mentorship of experts. Let the machinery do the credentialing as efficiently and effectively as AI can using digital platforms. Let the faculty and students figure out the learning without all the bureaucratic apparatus, standardization, and government control. So, I say CSU should optimize all the other steps Robbins lays out, but give the faculty and the students steps 4 and 6.
I'll try to do justice to this response with a longer piece soon. But again, I appreciate the willingness to think systematically about what CSU's embrace of AI could do for general education.