Jack & Jill reports 88% lower candidate screening cost with Jev
Read the provider's reported screening results, quality comparison and the limits of projected annual savings.
A focused replacement in a larger workflow
In TypeSafe's Jack & Jill case study, Jev replaces Gemini 3.1 Flash Lite for initial candidate scoring. A stronger model still reviews the shortlist. The evaluation used 150 live roles and tracked whether candidates requested by hiring managers remained on that shortlist.
Reported results
| Measure | Gemini baseline | Jev |
|---|---|---|
| Cost per 1,000 candidates scored | $0.755 | $0.092 |
| Median scoring time per role | 20.3 s | 10.3 s |
| Requested candidates retained | 93.9% | 94.6% |
TypeSafe reports 88% lower cost and roughly half the completion time. The 0.7 percentage-point quality difference was not statistically significant. The headline $500,000 annual saving is a projection including expected growth; the current-volume estimate is $265,000.
These are provider-reported results for one workflow, not jevai.me measurements or universal performance guarantees. The reusable pattern is a narrow scoring stage followed by deeper review. Test the same candidates, costs, recall and end-to-end timing in your own application before migrating.
Reviewed October 16, 2026; this is our review date, not a verified case-study publication date. Start with the integration guide.