AI candidate screening with a fit score you can actually defend.
Instalent scores each candidate against the criteria you set and shows its work - a breakdown and the evidence from their real work history, education and public code. Not keyword matching, and not a black-box number.
AI candidate scoring rates how well a candidate fits a specific role by analyzing their actual history against your stated criteria, returning a score with the evidence and reasoning behind it - not a keyword-match count.
Instalent reads each candidate's real profile - work history, education and public code on GitHub - and scores it against the criteria you define for the role. Every score comes with a breakdown showing which criteria were met and the evidence behind each judgment, so you can trust the ranking and defend it to a hiring manager. It classifies the companies in a candidate's background too, so 'senior at a startup' and 'senior at a large enterprise' don't read as the same line.
What you get
Every profile, scored the way you'd score it.
Score against your criteria
Define what matters for this role - skills, seniority, industry, company type - and Instalent scores every candidate against exactly that, not a generic template.
Evidence, not keyword matching
Each score points to the evidence behind it, drawn from the candidate's real history. You see why someone ranked where they did instead of trusting a keyword tally.
A breakdown you can defend
Every result shows which criteria were met and which weren't, so you can explain a shortlist to a hiring manager without re-reading every profile yourself.
Deep profile analysis
Instalent reads work history, education, and public code and contributions on GitHub - so an engineer is judged on what they've actually built, not just what a résumé claims.
Company context
Instalent classifies the companies in a background - startup, large enterprise, and the range between - so five years at a scrappy startup reads differently from five years at a giant.
Score at pipeline scale
Score a whole search, not one profile at a time, and sort the pipeline by fit so the strongest candidates surface first.
How it works
Automated candidate screening, from raw profiles to a reasoned shortlist.
- 01
Set your criteria
Tell Instalent what a strong fit looks like for this role - the must-haves and the nice-to-haves.
- 02
Instalent analyzes each profile
It reads work history, education and public code, and classifies the companies in the background.
- 03
Get a score with evidence
Each candidate comes back with a fit score, a breakdown of the criteria, and the evidence behind each judgment.
- 04
Sort and shortlist
Rank the pipeline by fit, review the reasoning, and move the strongest candidates into outreach.
Why it matters
Rank on substance, not on wording.
- Defend a shortlist with evidence instead of a gut call a hiring manager can push back on.
- Catch strong candidates a keyword filter would drop for using different words.
- Judge engineers on real, public work rather than résumé phrasing.
- Read seniority in context, because a title means something different at a startup and at an enterprise.
- Rank a whole pipeline by fit so review time goes to the people most likely to land.
FAQ
Frequently asked questions
Is this candidate screening software, and how does AI candidate screening work here?
It is screening in the sense that it puts a long list in a defensible order before you spend time on it - but automated candidate screening here never rejects anyone on its own. Instalent scores each person against your criteria with a written breakdown and the evidence from their real history, so you can see why a score landed where it did and overrule it. That is the difference between a candidate screening tool that filters people out silently and one that hands a recruiter a starting order they can defend.
Does AI candidate scoring create EU AI Act problems?
It is worth knowing where the line sits. The AI Act lists evaluating candidates as high-risk, and the obligations that come with that now apply from 2 December 2027 rather than the August 2026 date most guidance still quotes. Two duties already apply today though: the Article 50 transparency rules, and the Article 4 expectation that people using AI systems are trained to understand them. Instalent is built so the awkward part is not awkward - nobody is filtered out automatically, every score carries its reasoning and evidence, and a person makes the decision. Whether your specific deployment is high-risk is a legal question for your counsel, not a claim we will make for you; our trust page sets out what we do and do not assert.
How is this different from keyword matching?
Keyword matching counts whether words appear. Instalent analyzes a candidate's actual history against your criteria and returns a score with the evidence and reasoning behind it, so a strong fit who phrases things differently isn't dropped.
Can I set my own criteria?
Yes. You define what a strong fit looks like for the role, and every candidate is scored against exactly that rather than a fixed template.
What does it analyze?
Work history, education, and public code and contributions on GitHub, plus a classification of the companies in the candidate's background.
Why does a score come with a breakdown?
So you can trust and defend the ranking. Each score shows which criteria were met and the evidence behind each judgment, rather than handing you an unexplained number.
Can it score a whole search at once?
Yes. Score an entire pipeline and sort by fit so the strongest candidates surface first, instead of reviewing one profile at a time.
Can I try it for free?
Yes - start with a 7-day free trial and score a real shortlist against your own criteria.
Explore the rest of the platform
Explore
AI candidate sourcing
Describe a role in plain English and find the candidates you then score for fit.
Explore
Contact enrichment
Reveal verified email and phone for the candidates that score well, in the same flow.
Explore
For executive search
How search firms qualify a small, high-stakes shortlist with evidence they can present.
Explore
Interview questions to ask candidates
Scoring anchors written before the interview - the same discipline the fit score applies.
Explore
EU AI Act, GDPR & responsible AI
What the AI Act asks of hiring tools, which duties apply today, and what we will not claim.
Put your agent to work tonight
Start free and run your first search, enrichment and outreach in one flow.