University AI Builder Program
Students direct current AI agents to solve an approved problem, verify the work, implement only with permission, and turn the result into a portfolio they can explain.
Institution-owned and campus-published proof
From guest lecture to working curriculum.
The public record shows both ends of the continuum: a documented campus lecture at Oklahoma Christian and a six-session Johns Hopkins course built around collaboration with local businesses.
Applied Digital Marketing: Collaborations with Local Businesses
Instructor: Dennis Yu
Format: Six virtual Zoom sessions
Dates: March 27-May 1, 2025
The institution’s archived description says students work with local businesses to define objectives, create digital content, and reach target audiences.
A campus visit documented by the program and campus publication.
Eagle PR records Dennis Yu as a February 3-4, 2020 guest speaker. The Talon campus story separately documents his classroom lecture at Oklahoma Christian.
The two records preserve the campus context, dates, audience, and the business and communication topics taught during the visit.
Two useful entry points: Oklahoma Christian shows the focused guest-lecture format; Johns Hopkins shows how the same applied approach can extend across a six-session course.
A format that can grow with faculty demand
Start with one useful hour. Continue only when the work earns it.
Every format ends in a visible artifact and a student teach-back. Tool names may change; accountable scope, permissions, evaluation, implementation, and documentation stay durable.
60-90 minutes
One bounded workflow, built live with students rather than described from slides.
- Define the owner and acceptance test
- Direct an agent through one useful task
- Review the artifact and teach it back
Scope to teach-back
A repeatable sequence with public precedent in the Johns Hopkins six-session course record.
- Scope and source
- Build and verify
- Implement with approval and explain
10-12 weeks
A faculty-integrated studio covering agent systems, evaluation, governance, real implementation, portfolio, and handoff.
- Weekly accepted deliverables
- Client, campus, community, or labeled simulation
- Portfolio defense and reproducibility pack
The human + agent multiplier
Humans own the consequential decisions. Agents execute repeatable task steps.
Students keep responsibility for the work. They provide identity and context, define the scope, authorize action, inspect the evidence, and own the explanation.
- Create and secure personal or institution-bound accounts; accept terms and enable MFA.
- Supply firsthand context, interviews, objectives, consent, and the acceptance standard.
- Approve each live send, publish, purchase, credential grant, deletion, or production change.
- Resolve ethical questions, own errors, defend the contribution, and teach the result back.
- Research, extract, reconcile, classify, cite, and maintain a source and process ledger.
- Draft, build, compare, test, evaluate, repurpose, and prepare implementation artifacts.
- Run read-only canaries, accessibility checks, adversarial cases, and regression tests.
- Measure, document, prepare handoffs, and maintain a linked action and results log.
Student setup
What each student needs to build.
A phone is useful for capture, authentication, and review. The labs require a computer, student-owned identity, one paid AI plan, and only the project access the work actually needs.
Phone: capture, authenticate, review. Computer: build, test, run agents, manage files, and use version control.
University to private sector
One approved problem becomes proof a reviewer can inspect.
Students can use the same accountable project to earn experience, compete for a job, win a first client, or explore both paths. The artifact and the student’s explanation matter more than a promise about the destination.
Client path
Solve a bounded problem, deliver it cleanly, document the outcome, and earn the next introduction.
Job path
Turn the same work into a portfolio, oral defense, interview story, and role-ready evidence.
Both paths
Client work strengthens the resume; employment can sharpen the skills that win future clients.
What students carry forward: a working artifact, the tests and documentation behind it, a concise teaching story, and a repeatable process they can use in a job, with a client, or on their next project.
Human proof gallery
People building, teaching, and reviewing the work.
These images show the environment the curriculum is designed for: hands-on instruction, working agents, faculty-practitioner collaboration, and young builders teaching forward.
Watch the work
Full source recordings, not autoplay decoration.
Each card opens its original public recording. The page uses no embedded players, so visitors choose what to watch and the page remains fast and scannable.
The university offer
Bring a challenge. Leave with a working portfolio.
Start with a guest lab, a six-session sprint, or a 10-12 week agent studio inside an existing course.
The institution supplies the cohort, faculty liaison, room or Zoom environment, academic rules, and approved project context. Travel, venue, and student software are scoped separately.
The standard AI implementation workshop fee applies to conferences and other organizations. It is distinct from the qualified-institution offer.