Coming-soon platform
AI should clarify the decision—not hide how it was made.
This notice describes the intended design for AI-assisted degree-path predictions. Features may change before release; the active product will identify its current model, data sources, and limitations.
AI will not admit or reject a student, award or deny transfer credit, determine financial aid, confer a license, guarantee employment, or replace the Degree Dean’s review and the school’s official decision.
1. Intended uses
- Extract courses and potential academic assets from user-reviewed transcript data.
- Identify possible equivalencies, alternative-credit opportunities, and unresolved requirements.
- Estimate completion-time and total-cost ranges under stated assumptions.
- Compare fastest, lowest-cost, and best-fit scenarios.
- Highlight uncertainty, outdated source data, conflicts, and required human verification.
- Suggest one next best action based on the current plan state.
2. Inputs and sources
Potential inputs include user-entered goals, prior credits, certifications, professional training, time and budget preferences, current program rules, transfer guides, public equivalency information, and human-reviewed institutional data. Source dates and relevant assumptions should be visible with output.
3. Output design
Predictions will be expressed as ranges or ranked scenarios—not certainties. Where feasible, the interface will display confidence, supporting factors, missing evidence, source freshness, and the impact of changing pace, budget, or institution. “Modeled” means calculated from available data and assumptions; it does not mean approved by a school.
4. Human review and contestability
Users can flag an error, correct inputs, request human review, ask for an explanation, object to optional AI processing, and seek a non-AI route where reasonably available. A coach reviews recommendations before they become part of a client deliverable. Official decisions must come from the relevant institution or authority.
5. Data protection
Transcript parsing is intended to be local-first when practical, with visible review before selective persistence. Identifiable transcripts will not be used to train public AI models. Data sent to a contracted AI provider will be minimized to the defined task, subject to vendor review and contractual safeguards, and retained according to the disclosed lifecycle.
6. Quality, bias, and monitoring
The platform will be evaluated for extraction errors, inconsistent pathway rankings, source staleness, disparate performance, overconfidence, and harmful recommendations. Material issues may lead to correction, feature restriction, source removal, model replacement, or suspension. Users should report suspected errors through the contact or privacy-request workflow.
7. Professional and regulated outcomes
Career and degree information is educational. It is not a professional-license determination, employment screening report, credit decision, legal opinion, or financial-aid award. Psychology and healthcare pathways do not independently authorize clinical practice; criminal-justice pathways do not confer law-enforcement authority.
8. Change control
Before material AI changes, The Degree Agency should update its data map, risk assessment, use-case approval, vendor review, testing evidence, user notice, retention schedule, and incident playbook. The release date and current notice version will be shown in the live platform.