Optimises for vendor demo, not real hiring.
Full-automation AI tools demo brilliantly and break in production. The verdicts the model gets right are easy; the ones it gets wrong are the ones that matter, and full automation hides them.
Most AI resume screening tools replace the recruiter. ShortlistTable removes the keystrokes and leaves the judgement intact — per-criterion AI verdicts, source citations, one-click override, exports that land in your ATS clean. The screening pass without the screening grind.
| Candidate | Must-have 1 | Must-have 2 | Must-have 3 | Priority | |
|---|---|---|---|---|---|
Alice Chen Ledger · 7y | Yes | Yes | Yes | Call first | 1st |
Marco Silva CloudCo · 12y | Review | Yes | Review | Review | 2nd |
Nina Patel Indie · 4y | No | No | Review | Hold | 3rd |
The first wave of AI recruiting tools promised full automation: AI screens, AI ranks, AI rejects, AI schedules. The market noticed pretty quickly that the “AI rejects” step was the one creating lawsuits and PR incidents — and the “AI ranks” step was the one producing nonsense.
We took the recruiter-first design instead. The AI handles the mechanical parts of screening — reading every resume against every criterion, citing evidence, queuing ambiguous cases. The recruiter keeps the judgement. Same throughput, less risk, better hires.
Full-automation AI tools demo brilliantly and break in production. The verdicts the model gets right are easy; the ones it gets wrong are the ones that matter, and full automation hides them.
Some AI tools just summarise what your recruiters already did. Useful for managers, not for the recruiter doing the actual screening at 8am on Monday.
The right design: AI handles the per-criterion read, the source citation, the keystroke-heavy work. Recruiter handles the judgement, the override, the rejection. Same throughput; better defensibility.
200 resumes × 5 columns = 1,000 cells the AI populates with verdicts and source citations. The recruiter doesn’t have to do the 1,000 reads.
Every verdict carries the exact resume sentence behind it. If the citation is wrong, you adjust the column wording. If the verdict is wrong, you override the cell.
Below your confidence threshold, the AI puts the cell in a review queue instead of guessing. You spend time on the cases that actually need a human.
Disagree? Click, override, optional note. The AI verdict stays in the audit history; your override becomes the active value across the workspace and every export.
CSV / XLSX presets for Greenhouse, Ashby, Lever, Workable, Bullhorn, JobAdder — AI verdicts and evidence columns survive the round-trip.
Low-fit candidates land in a hold queue with evidence intact. No silent screen-outs, no audit blind spots.
| Property | Full-automation AI | ATS AI scoring | Manual screening | ShortlistTable |
|---|---|---|---|---|
| Recruiter is final decision-maker | ✕AI decides | –Limited override | ✓Yes | ✓Every cell overridable |
| Per-cell evidence | ✕Score only | ✕Rank only | ✓Manual notes | ✓Sentence + page |
| No auto-reject | ✕Common | ✕Common | ✓Manual | ✓Hold queue only |
| Time-to-shortlist (200 resumes) | ✓Fast (risky) | –Variable | ✕Days | ✓Hours (safe) |
| ATS-shaped export | –Generic | ✓Native | ✕Manual | ✓6 presets |
No. The AI replaces the read-the-resume step. The screening call is still where you read the candidate as a person, ask sharp follow-ups, and form your own read on fit. The AI gives you back the time to spend on those calls instead of on the resume pile.
Override the cell in one click. The AI verdict is logged; your override becomes the active value. The audit trail captures both.
Yes — CSV / XLSX presets shaped for Greenhouse, Ashby, Lever, Workable, Bullhorn, and JobAdder. Drop the export into your importer; verdicts and evidence both survive.
PDF, DOCX, TXT, and pasted text. The free browser tools (JD↔CV matcher, missing requirements finder) parse TXT only; the paid product handles PDF and DOCX at batch scale.
Per-workspace plans with a monthly resume volume. The free tier covers 25 resumes/month so you can validate the workflow before committing. Pricing details on the pricing page.
Yes — the free browser tools at /tools (JD↔CV matcher, JD critique, bulk resume parser, etc.) all run locally with no account. Use them to test the matching logic on a single candidate or a small batch.
Drop 25 resumes, write 5 columns, get a sheet with per-cell evidence. Free, no credit card.