A four-year public university receiving transcripts from 140 feeder institutions, with admissions evaluators re-keying course lists and mapping equivalencies in a shared spreadsheet. We built the normalisation layer: LlamaParse on typed transcripts, LandingAI on signed and scanned copies, every course mapped against the local catalogue with under-threshold equivalencies flagged for evaluator review.
| Applicant | Source course | Hrs | Equivalency | Conf. | Status |
|---|---|---|---|---|---|
| T0094213 | ENG 1020 · Prairie Ridge CC | 3.0 | ENGL 101 | 0.97 | posted |
| T0094214 | HIS 2210 · Lakeshore State | 3.0 | HIST 206 | 0.94 | posted |
| T0094215 | BUS 3005 · Cedar Valley CC | 3.0 | BUSN elective | 0.81 | dept review |
| T0094216 | MATH 1410 · Northpoint CC | 4.0 | MATH 165 | 0.96 | posted |
| T0094217 | PSYC 201 · Riverside University | 3.0 | PSYC 111 | 0.93 | posted |
| T0094218 | ARQ 4408 · Universidad de los Andes | 4.5 | no equivalency | 0.42 | faculty |
| T0094219 | BIO 1500 · Greatlakes Tech | 4.0 | BIOL 150 | 0.95 | posted |
| T0094220 | SOC 210 · Westbrook College | 3.0 | SOC general elective | 0.78 | dept review |
| T0094221 | COMM 1100 · Prairie Ridge CC | 3.0 | COMM 110 | 0.97 | posted |
At a glance
One admissions office, 140 feeder institutions, one course catalogue. The equivalency match was the piece evaluators ran out of hours for in peak season.
The engagement
The stack
ISO 27001 · ISO 9001 · FERPA scope · DPA and NDA signed at kickoff.
Before, the transfer office
Evaluators opened every transcript. They typed the course list into a spreadsheet, ran the equivalency in their head or against the binder, and posted the results to Banner. The pattern worked, and then peak season arrived.
A transcript arrived with 28 courses. The evaluator typed each course into the spreadsheet, mapped it to the local catalogue, and marked the equivalency. On a clear transcript, 40 minutes. On a transcript from an overseas institution, 80 or more.
Pre-build baseline: 40 to 80 minutes per transcript on re-key and equivalency.
The binder mapped courses from the 20 largest feeder institutions. The senior evaluators knew the next 60. The remaining 60 feeders got a best-guess match, which the evaluator flagged for review, which often came back to them anyway.
Pre-build baseline: equivalency consistency varied outside the top 80 feeders.
Spring applications hit, transfer transcripts arrived, and the window between receipt and advising was shorter than the evaluator's re-key time allowed. Overtime covered some of it. Slippage into advising week covered the rest.
Pre-build baseline: approximately 18% of transcripts evaluated after the advising-week cutoff.
What we built
The pipeline follows the same five stages we run on every normalisation engagement. The course-catalogue mapping and the equivalency rule library are tuned against this university's active catalogue.
Parchment webhook, Slate upload listener, applicant portal form, paper scanner drop to SFTP. Every transcript assigned a single applicant ID tied to the admissions case.
Transcript type tagged on ingest. Official transcripts routed directly, unofficial flagged for later verification, overseas transcripts routed through the international evaluation step. Classification confidence below 0.90 holds the document.
Feeder institution, course codes, titles, grades, credit hours, term. LlamaParse on typed transcripts, LandingAI on signed and scanned copies.
Each course mapped against the local catalogue through the equivalency rule library. Direct match scored. Under-threshold equivalencies flagged for evaluator review. Below 0.85 confidence, the course holds.
Clean equivalencies posted to Banner against the applicant record, with the source transcript page attached per course. Held courses route to a named evaluator queue with the reason in plain English.
After, the numbers the transfer office signs off
Same evaluators, same catalogue, same 140 feeder institutions. The pipeline parsed every transcript, ran the equivalency, and routed ambiguous courses to the evaluator. Peak season stopped overflowing the advising-week cutoff.
Evaluators still own the ambiguous course. They still write every new equivalency rule into the library. The difference is that peak season finishes inside the advising window, and the spreadsheet is not the system of record anymore.
From the desk
Peak season used to mean overtime and slippage. We closed on time this spring for the first time I remember.
Admissions evaluation leadFour-year university, Great Plains
Handover
The engagement ends at a clean handover. The transfer office runs the pipeline; Hexaa stays on call for a fixed retention period, then steps back.
Related cases
Each links to a named client, a named document, and the system the clean data lands in. We publish only what the client signed off to publish.
NIH and NSF packets assembled against the funder template, with section ownership named and missing items surfaced for pre-award.
→State and local government · 2026Mid-sized city · permit intake, one queue per typeBuilding, zoning, and right-of-way permits classified into one queue per type with attachment checks at intake before the clerk opens the file.
→Construction · 2026General contractor · AIA G702 and G703 packet assemblyAIA pay application packets assembled with supporting attachments named per line. Procore as the system of record.
→Free 30-minute call
You'll leave with a clear next step.
A transcript arrives with ENG 1020 at a feeder community college. The pipeline matches it against the local catalogue through the equivalency rule library, posts the direct match to Banner, and flags the one course where the rule library has no rule yet. Evaluators see ambiguity, not data entry.
