Find My Path AI Agent System

A student asking “Which school fits me?” is not asking for a search result. They are asking someone to weigh academic readiness, budget, destination, programme preference and personal goals—and then explain the trade-off.

From questionnaire to a defensible shortlist

Find My Path turns that conversation into a staged workflow. The profile collector first resolves missing information instead of guessing. The school finder then queries PathCAN’s own Go APIs, scores realistic candidates and returns three options that can be explained. A human-approval step keeps the student in control before the system creates a formal recommendation.

The important design decision was to keep these responsibilities separate. Retrieval can be improved without rewriting the interview. Ranking rules can change without touching PDF generation. A weak result can be traced back to the profile, the source data or the scoring step instead of disappearing inside one long prompt.

What the student receives

The service supports normal JSON responses and SSE streaming for the conversational interface. Once a shortlist is approved, it can generate a branded PDF with school details and QR codes and hand it back through the Go platform for delivery.

This makes the agent useful beyond the demo moment: the conversation becomes an inspectable recommendation that a student can review, share and act on.