Must-haves
The requirements that genuinely rule someone out. Keep this list short and mean it.
The product
A shortlist, and the reasons for it.
Candidate screening software for one job: taking a batch of applicants and giving you back the people worth reviewing, with the evidence that put them there.
People worth a closer look
- 194
Candidate 0413
- 7 yrs in role
- Ships in Rust
- Team lead
- 291
Candidate 0088
- 6 yrs in role
- Open-source
- Mentors
- 387
Candidate 1102
- 5 yrs in role
- Scaled infra
- 484
Candidate 0641
- 4 yrs in role
- Domain match
The flow
What actually happens.
No pipeline to configure. No stages to design. Four steps from pile to shortlist.
Create the role
Describe the job in normal language. This is the context every later decision is measured against.
Set the criteria
Must-haves, important experience, nice-to-haves, disqualifiers, and constraints. You are defining what matters, not writing a keyword list.
Add the batch
Upload resumes or applicant data in bulk. Hundreds or thousands at a time, from an inbox, a folder, a spreadsheet, or a job board export.
Review the shortlist
Prioritised candidates, each with what they met, what supports it, and what is missing. Open anyone who was not shortlisted at any time.
Criteria
Screening is only as good as what you asked for.
Keyword matching is why good candidates get thrown away and why weak ones get through. You set the requirements, and each one is evaluated on the evidence in the application rather than on whether a phrase appeared.
Important experience
What you would strongly prefer. Weighed and reported, not treated as a hard gate.
Nice-to-haves
Useful signals that break ties without deciding the outcome on their own.
Disqualifiers and constraints
Location, work authorisation, licences, and anything else that makes a hire impossible.
What you get back
Every candidate arrives with their reasoning attached.
A shortlist is only useful if you can act on it without redoing the work. So each surfaced candidate shows the requirements they met, the specific evidence in their application, and the requirements no evidence was found for.
You can disagree with any of it. That is the point of showing it.
- Requirements met, with the supporting evidence
- Areas of strength worth asking about
- Missing evidence, stated rather than assumed
- Applications flagged as needing a human look
- 194
Candidate 0413
- 7 yrs in role
- Ships in Rust
- Team lead
- 291
Candidate 0088
- 6 yrs in role
- Open-source
- Mentors
- 387
Candidate 1102
- 5 yrs in role
- Scaled infra
- 484
Candidate 0641
- 4 yrs in role
- Domain match
- 581
Candidate 0297
- 5 yrs in role
- Regulated industry
No ATS required
Applications come from wherever they came from.
This is not an integration project and it is not an ATS. Nothing needs to be connected before you can screen a batch, and nothing you already use has to be replaced.
- Inbox
- Spreadsheet
- Job board
- Drive folder
- Referrals
SortTheStack
No ATS required
- Ranked shortlist
- Evidence per candidate
- Gaps to probe
How we treat uncertainty
Efficiency should not mean throwing good people away.
The failure mode of automated screening is not that it is slow. It is that it silently removes someone you would have hired, and you never find out.
Role-specific criteria
Every evaluation is against the requirements you set for this role. There is no universal quality score being applied behind the scenes.
Transparent reasoning
The evidence behind every result is visible, so you can check the reasoning rather than trust it.
Uncertainty surfaced
Ambiguous applications are marked as needing human judgement instead of being sorted as if the answer were clear.
People make the decisions
No rejection is sent unless someone on your team sends it. Applicants are only set aside automatically by rules you choose, and you can override them. The software narrows the reading, not the choice.
No predictions about performance
This evaluates evidence in an application against your stated requirements. It does not claim to know how someone will do in the job.
Consistent criteria, human judgement
Structured screening makes your requirements explicit and applies them the same way to every applicant. The hiring decision stays with you.
Questions
How it works, specifically.
How does candidate ranking work?
You define the criteria for the role: must-haves, important experience, nice-to-haves, and disqualifiers. Each applicant is evaluated against those criteria, and the ones with the strongest supporting evidence surface first. The criteria are yours, so the ordering is explainable rather than a mystery number.
Can I see why a candidate was recommended?
Yes. That is the point. Every surfaced candidate shows which requirements were met, the evidence in their application that supports it, and what could not be found. If you disagree with the reasoning, you can see exactly where it came from.
What happens if the system is not sure about a candidate?
It says so. Ambiguous applications are flagged as needing a human look rather than quietly sorted into a pile you will never open. A screening tool that hides its uncertainty is how you lose the person you should have hired.
How many applications can I screen?
Batches in the hundreds or thousands are the normal case, not the edge case. Screening capacity by plan is on the pricing page.
Can I upload resumes in bulk?
Yes. Add the whole batch at once rather than one file at a time. That is the entire reason this exists.
Can recruiters use it for high-volume hiring?
Yes. High-volume roles are where the time saved is largest, because the first pass is where most of the hours go. Agency recruiters and internal teams both use it the same way: screen the batch, review the shortlist, spend the rest of the day talking to people.
Bring us the pile.
Create a role, add your applicants, and see what a shortlist with its reasoning attached looks like.