Problem clarity
The user, need, success criteria, constraints, and scope are understandable.
The program operating system
A project-led path from idea to evidence.
Every stage produces something reviewable. The sequence reduces confusion, keeps scope visible, and makes the final result easier to explain.
Six build stages
The same discipline, adjusted to program depth.
A seven-day prototype and a ninety-day product require different depth, but both benefit from clear scope, visible work, and an honest release.
Problem framing
Define the user, workflow, pain point, success criteria, constraints, risks, and non-goals before selecting features.
Architecture and roadmap
Plan the data, model behavior, interface, evaluation, tools, deployment, and a sequence of achievable milestones.
Implementation
Build the smallest useful workflow first, then add capability in reviewable increments with visible version history.
Review and iteration
Use feedback to improve decisions, code structure, user experience, testing, failure handling, and documentation.
Validation and release
Test expected behavior and edge cases, deploy the project, and state limitations instead of hiding them.
Demonstration and narrative
Explain the problem, architecture, evidence, trade-offs, result, and what you would change with more time.
Student responsibilities
You own the work.
- Protect regular build time
- Share blockers before they become delays
- Respond to feedback with visible changes
- Test and document honestly
- Present your own decisions and limitations
Program responsibilities
We make the next decision clearer.
- Provide a structured sequence
- Review relevant work and decisions
- Clarify priorities when scope expands
- Keep feedback concrete and actionable
- Maintain professional learning boundaries
The project quality rubric
What strong proof of work should show.
This rubric guides feedback and project quality. The scheduled tenure certificate records participation in the program period; it does not assign a grade or certify mastery of every criterion.
Working behavior
The main workflow runs and failure cases are not ignored.
Technical reasoning
Architecture and implementation choices can be explained.
Evidence and communication
Repository, documentation, deployment, and demonstration support the story.
Applied AI build programs
View program detailsReady to build evidence, not just complete content?
Choose the program depth that fits your current experience and the project you want to ship.