Degree planning that survives reality.
Seyir turns a student’s goals and institution-controlled academic records into a complete semester plan—then verifies every hard rule before that plan can be shown.
AI proposes. Rules verify.
“Keep next spring lighter without delaying my path.”
Built for a harder standard
A plan that sounds plausible is not enough.
One planning flow.
Two kinds of intelligence.
Generative AI can reason across goals and tradeoffs. Academic rules demand exactness. Seyir gives each the job it is suited to do—and refuses to blur the boundary.
Student goals
Plain-language preferences for pace, workload, timing, and priorities.
- “Make spring lighter”
- “Protect my graduation date”
Institution-controlled records
A governed snapshot of the facts the plan must respect.
- Program and catalog rules
- Course history and exceptions
Author a semester-by-semester path
Check the plan against academic rules
Invalid after repair? The schedule is withheld—not softened into a suggestion.
From intent to a plan that can be checked.
A single request becomes an auditable planning result in three deliberate moves.
-
01
Freeze the facts
Seyir starts from an authorized, versioned snapshot of the student’s program, academic history, catalog, offerings, and approved exceptions.
Institutional truth stays in control. -
02
Author the full path
The planning model interprets the student’s goal and proposes every remaining semester—not just the next handful of courses.
Tradeoffs are considered across the degree. -
03
Prove or withhold
Deterministic code evaluates hard constraints. One structured repair is allowed; if the plan still fails, no schedule is returned.
A failed check never becomes a student-facing plan.
A clearer plan for every person around it.
Seyir is being developed as shared planning infrastructure: useful to students, reviewable by advisors, and governable by institutions.
For students
Turn a real-life goal into a complete path—and see workload, credits, and progress in context.
- Ask in plain language
- See every remaining semester
- Understand tradeoffs earlier
For advisors
Start the conversation with a structured, checked proposal instead of an empty worksheet.
- Review clear planning assumptions
- Spot warnings and exceptions
- Spend time on judgment, not assembly
For universities
Explore an AI planning experience without handing academic authority to a language model.
- Keep source records governed
- Pin versioned academic rules
- Fail closed when facts are incomplete
The model is the author.
It is never the authority.
A language model can assemble and revise the path. The institution’s published facts and Seyir’s deterministic evaluator decide whether that path is valid enough to display.
- Authorized inputs only
The browser does not invent transcript or degree facts for the planning authority.
- Immutable planning snapshot
Each result is checked against one pinned set of academic facts.
- Fail-closed output
Unknown, stale, or contradictory rules stop the plan instead of being guessed through.
- Course identityNo unknown or duplicate coursesPASS
- Requisite sequencePrerequisites and corequisites satisfiedPASS
- Term availabilityKnown offerings respectedPASS
- Degree coverageRequirement totals and allocations completePASS
If any required check remains invalid:
PLAN WITHHELD · NO SCHEDULE RETURNED
Seyir supports planning and advising conversations. It is not an official degree audit, enrollment decision, course registration system, or substitute for institutional approval.
Build the planning layer your students can actually reason with.
Seyir is seeking thoughtful pilot conversations with institutions that want to explore verified, student-centered degree planning—without overstating what AI should decide.
Institutional data fitCatalog, program rules, student records, and versioning.
Planning experienceStudent goals, advisor review, and complete-path clarity.
Verification behaviorRule coverage, failure modes, provenance, and handoffs.