Credit Cruncher & TransferMe
Working out which transfer credits will count meant reading transcripts by hand, one course at a time — slow for staff and opaque for students.
- AI and automation
- Internal tools
- Quality assurance
- Product Operations
AI-assisted step Human decision Observed friction
Summary
- Challenge
- Speed up transfer-credit evaluation for enrollment staff, then extend it to students, without losing accuracy or trust.
- Role
- I supported the product team: QA testing, both product walkthroughs, TransferMe's user stories and success metrics, design sessions for the student-facing version, and a prototype using real articulation data.
- Team
- Product manager, my manager, engineering, design, and enrollment specialists
- Duration
- 2026
- Methods
- QA testing, product walkthroughs, user stories and acceptance criteria, success metrics, design sessions, prototyping
- Outcome
- Credit Cruncher is used by enrollment specialists as a pilot; TransferMe is its student-facing version. I don't own either product.
Context and problem
Prospective transfer students want to know which of their credits will count. Answering that meant an enrollment specialist reading an unofficial transcript and looking up each course in a transfer-credit guide, one at a time.
Credit Cruncher uses AI-powered document reading to pull courses from transcripts and match them to ASU equivalents from the existing transfer-credit database, so specialists can evaluate a transcript far faster. TransferMe brings the same capability to students directly.
My role and collaborators
- What I did
- QA testing of new features; the product walkthroughs for both tools; TransferMe's user stories and success metrics; participation in design sessions for the student-facing version; and a prototype for one transfer pathway using real articulation data.
- What the team owns
- A product manager owns Credit Cruncher and TransferMe. Engineering built the tools; design owns the experience; enrollment specialists use the tool and own every credit decision.
How the tool works
- Step 01
Upload transcripts
Several at once, or add courses by hand
- Step 02 AI-assisted
AI reads them
Courses, grades, credits extracted
- Step 03 Human decision
Review & edit
Bulk edit, remove rows, fix grades or institution
- Step 04
Find equivalencies
Matched from the transfer-credit database
- Step 05
Change institution
Without starting over
- Step 06
Generate report
Credits, courses needing evaluation, disclaimer
Product decisions I saw up close
- Sign-in for uploads
- Uploading a transcript requires signing in, to protect students' records.
- Honest results
- When fewer than half of a student's credits transfer, the result is shown as a plain list rather than a celebratory reveal, to avoid a misleading moment.
- Features before savings
- While the tool was still a pilot with a small group of users, we presented features and workflow rather than time-savings claims, because the timing data wasn't yet a fair comparison.
QA and documentation
- Tested new features in the QA environment before release, including bulk editing and the change-institution flow.
- Ran an end-to-end flow check and reported sign-in lag, a "Sign in" label that didn't switch to "Sign out", a logo link that went to the wrong place, and broken navigation back into the app.
- Built the stakeholder product walkthrough as a six-step deck — student details, upload, review imported courses, review equivalencies, change institution, generate report — using focused screenshots from QA. It became the template for the team's later walkthroughs.
- Wrote the TransferMe product walkthrough on the same model, covering the student's path from entry and sign-in through onboarding, adding coursework, the transfer summary, and feedback. I also contributed to its launch support and escalation documentation.
- Wrote TransferMe's user stories: 12 epics and 27 stories covering the whole flow except authentication, each with acceptance criteria and linked to its design.
- Defined TransferMe's success metrics, each with a target: satisfaction score, funnel completion, transcript-reading success, conversion to a next step, and time from starting an evaluation to reviewing it.
“You did a great job with Credit Cruncher.”
Design sessions and prototype
For the student-facing version, I joined design sessions on the hardest cases — transcripts that list several institutions, unreadable files, and courses that need to be added by hand. Separately, I built a prototype for one transfer pathway: a student picks their community college and associate degree and sees which matching ASU concentration it maps to, using real articulation data.
What I learned
Good AI products are mostly good non-AI decisions: where the model stops, how results are framed, what happens when nothing matches, and who reviews the edge cases.
What I'd improve next
- Design a fair comparison of manual and assisted evaluation, with the same specialists doing both, before making time-savings claims.
- Watch how students use TransferMe after launch to learn how many have transcripts on hand versus adding courses manually.
My Path: AI Student Retention Agent
Online students often withdraw over small, fixable barriers — a hold, a small balance, a missing aid document — that no one spots before the drop date.
- My role
- Product owner and builder: problem framing, rules, AI guardrails, coach workflow, and delivery
- Evidence
- A working prototype on synthetic data: six tested barrier rules, AI drafts behind strict guardrails, and a coach who approves every message.
- AI and automation
- Product Operations
- Quality assurance
- Strategy