§ Suite · ServicesFive tools · one workspace
What Illuma actually does.
Each service is one instrument. Pick the tab to see the tool, how it works, and what makes it different.
§ Under the hoodWhat the model actually measures
You sit at the center; every dot is another applicant plotted by similarity. Closer dots = more similar profile (major + stats + hooks). Ranked from an extensive corpus of self-reported historical applicants.
Output includes
- Outcome tabs — Admitted / Waitlisted / Rejected, filter to what you need
- Full profile per card — stats, major, hooks, activities
- Cross-school outcomes visible on each match
- Weighted matching — major first, then GPA, then test score
- Server-side caps on distinct applicants to protect source privacy
Under the hood
ML models + calibration layers, not a ChatGPT wrapper.
1
Outcome-aware presentation
Each candidate is bucketed into admits, waitlists, or rejects at the target school — you toggle which story to see.
2
Cross-school context
Every card surfaces their full application shape (other admits + rejects) so you can read fit patterns.
3
Evidence-first framing
We show what happened, not what will happen. The corpus is self-reported and skews toward higher-stat applicants — surfaced everywhere it matters.
Ready to try it?
First eval free. No credit card required to sign up.