Working prototype · Agentic AI for university admissions

Every application, pre-screened

Placedor's AI agents read each file, check it against the programme's entry rules, collect results from the verification services universities already use, confirm what those services miss with the issuer, and sort the queue. A person approves every case. The university keeps every admission decision.

Agents Intake→ Extraction→ Verification→ Rules engine→ Integrity→ Triage→ Human sign-off + Programme-data agent

Pre-screened queue

Entry rules check · MSc Business Analytics (sample programme)

Rule	Applicant	Result

Documents and verification

Document	Extracted	Verified by	Result

Agent log

Approving confirms the file is complete and checked. The admission decision stays with the programme.

How the agents work

Seven agents, one reviewer

Each agent has one job and only the tools it needs. A person signs off every case, and the university makes every admission decision. A seventh agent watches official programme pages in the background and queues rule changes for staff approval.

Step 1Intake
Checks the uploads and asks the applicant for missing pages.
AI agent
Step 2Extraction
Reads transcripts, certificates and test reports into fields.
AI agent
Step 3Verification
Credential networks and test providers first, then the issuer.
AI agent
Step 4Rules engine
Coded entry rules: degree class, GPA conversion, English scores.
AI agent
Step 5Integrity
Flags signs of editing or mismatch. Never rejects on its own.
AI agent
Step 6Triage
Ready, needs information or needs review, with the reason.
AI agent
Step 7Sign-off
A reviewer approves the pre-screen. The programme decides.
Human reviewer

AI agentHuman involved