The real incumbent. Not a product — a recruiter, a stack of 200 applications, and a Sunday.
A human reading a résumé brings context no system has: they know the hiring manager, they recognise the company that candidate left, they can feel that a career move means something. On a pipeline of twelve applicants, a person reading carefully is the best screen there is.
| manual résumé screening | Talyn | |
|---|---|---|
| Time per role | Roughly 25–30 hours across 200 applicants, most of it on people who will be rejected. | Zero hours to run; you read a ranked shortlist. |
| What it can see | A document the candidate wrote about themselves. | What the candidate can actually do, watched live. |
| Who gets cut | Gaps, non-traditional paths, unfamiliar employers — cut to ration time, not because they were assessed. | Everyone gets interviewed; the saves are surfaced. |
| Consistency | Applicant 190 is not read the way applicant 4 was. | The same bar for the first and the last. |
| Defensibility | A rejection reason reconstructed from memory, weeks later. | The exchange that produced it, recorded. |
your pipeline is small enough that a person can read every application properly — under about thirty per role.
inbound has outgrown the hours available, and the honest answer is that most applications are not really being read.