Get Started

Screen every CV like it's the only one.

Score, rank, and compare every applicant against your real job requirements, with AI as your screening assistant, not your decision maker.

Book a Demo Call See Capabilities
🔒 Runs in your Azure tenant 🛡️ GDPR & EU AI Act aware ⚡ Evidence-traced rankings
"We get 400 applications for one role. By CV number 60, every reviewer is skimming. Two recruiters score the same candidate completely differently, and the best CV might be sitting at number 312."
5×
faster than manual
CV screening
100%
scores traced to
CV evidence
0
CVs skimmed, skipped
or forgotten
The Solution

A ranking engine built for real hiring.

Not a keyword filter. Not an ATS plugin. A structured, multi-stage reasoning system that reads every CV the way your best recruiter would on their best day: thoroughly, consistently, and with a complete evidence trail.

Define the role requirements in plain English: must-haves, nice-to-haves, and how much each one matters. The system scores every applicant against every requirement, ranks the full field, and shows the exact CV passage that justified each score. Your recruiters decide; the system makes sure nothing decisive was missed.

Natural-language requirements Weighted scoring Full traceability Azure-native Human decides

Architecture Overview

📄

Your Applications

CVs in PDF and Word, cover letters, and application forms, ingested from your ATS export, SharePoint, or inbox

🧠

Knowledge Layer

Azure AI Search · Azure AI Foundry · candidate profiles with hybrid vector + keyword retrieval

⚙️

Ranking Engine

Proprietary orchestration framework: multi-stage scoring against weighted, role-specific criteria

📊

Delivery

Azure Functions · Azure AD · Excel shortlist export · Structured review UI in your tenant

See It In Action

From 400 applications to a ranked shortlist in under an hour

Watch the CV Ranker score a real applicant pool against a Senior Data Engineer profile: 14 weighted criteria, 3 hard must-haves, every score traced to the exact CV passage that produced it.

Core Capabilities

Four capabilities. One candidate pool.

Every capability is built on the same underlying knowledge layer, so you can start with one vacancy and expand to every role without rebuilding the foundation.

🎯

Criteria-driven CV scoring

Define the role in plain English: must-haves, nice-to-haves, and weights. No schema, no coding, no vendor involvement. The system scores every applicant against every requirement and ranks the full field, with evidence for each score.

Natural-language requirements Must-have filters Weighted scoring Per-role change
🧩

Reads CVs, not keywords

The strongest candidates rarely use your exact vocabulary. The system understands that "Data Platform Lead, 2019–2024, Synapse & ADF" means five years of Azure data engineering, and that a buzzword-dense CV with no substance behind it doesn't.

Semantic skill matching Tenure calculation Title normalisation Gap & inconsistency flags
🔍

Ranked shortlists with full reasoning

Every candidate gets a score breakdown per criterion (Met / Partial / Missing) with the CV passage behind each one. Drill into any score, verify against the source, override if needed, and export the shortlist to Excel.

Met / Partial / Missing Evidence citation Human override Excel export
💬

Natural-language candidate Q&A

Beyond structured runs, recruiters ask ad-hoc questions across the entire pool: "Who has Azure certification and German C1?", "Which candidates led a team of 5+?", "Show everyone available within 3 months." Every answer cites its source CV.

Pool-wide queries Evidence-traced answers Aggregation queries Multi-language CVs
The Interface

Built for recruiters,
not engineers.

A rich review interface, not a black-box score, not a CV dump. Every ranking is filterable, drillable, and exportable. Built for HR teams and hiring managers who need to defend every shortlist decision.

  • ✓Rank the full field or filter by must-have requirements
  • ✓Click any score to see the CV passage and reasoning chain
  • ✓Compare candidates side by side, criterion by criterion
  • ✓Human override on any score: your team stays in control
  • ✓One-click export to Excel or your shortlist template
CV Ranker: Senior_Data_Engineer (247 applicants · 14 criteria)
Ranking: Top of 247 31 shortlisted · 164 partial · 52 must-have fails
#1 · 94 M. Keller: all 3 must-haves met, 9 of 11 weighted criteria
5+ yrs Azure data engineering: MET, "Data Platform Lead, 2019–2024" with Synapse, ADF and Databricks delivery. Team leadership: MET, built and led a team of 6 engineers. German C1: MET, stated native speaker, CV submitted in German.
Keller_CV.pdf · p.1–2 · Confidence: High
#2 · 89 A. Nowak: must-haves met, 1 weighted gap flagged
Power BI at scale: PARTIAL, dashboard delivery confirmed, but no evidence of enterprise rollout or governance. Recommend probing in first interview. All other criteria met with cited evidence.
Nowak_CV.pdf · p.2 + cover letter, p.1
#212 · N/A Must-have not met: 5+ yrs Azure data engineering
CV shows strong AWS background (Redshift, Glue) but Azure experience limited to a 6-month migration project in 2023. Excluded from shortlist by hard filter, not by a hidden score.
Candidate_0212.pdf · p.1 · Hard-filter exclusion, fully logged
Getting Started

Running on a real vacancy within one week

No migration. No new ATS. The system connects to your Azure environment and works with the CVs you already have.

1

Discovery workshop

We map your hiring workflow, role profiles, and screening criteria. You leave with a concrete pilot scope and proposal.

½ DAY · FIXED FEE
2

Pilot on a live vacancy

We deploy the system on a real applicant pool: your actual CVs, your actual requirements. Typically 2–3 weeks.

2–3 WEEKS
3

First live screening

Your recruiters run the first full ranking on a production vacancy. We support the review session and tune the criteria together.

DAY 1 OF PRODUCTION
4

Expand & evolve

Add role profiles, ATS integration, or adjacent use cases (talent-pool mining, internal mobility) on the same foundation.

OPTIONAL RETAINER
Where Keyword Filters Fail

The match a keyword filter misses, found with evidence.

Keyword filters reject the candidate who wrote "Data Platform Lead" instead of "Data Engineer", and rank up the one who pasted your job ad into their skills section. The system reasons across titles, dates, projects, and cover letters, so the ranking reflects what candidates actually did, not which words they used.

Titles ↔ Skills Dates ↔ Tenure CV ↔ Cover letter + multi-language pools

Live example: requirement matching

Requires: → 5+ years Azure data engineering experience
CV says: → "Data Platform Lead, 2019–2024, Synapse, ADF, Databricks"
Reasoning: → Role + stack = Azure data engineering · tenure = 5 years
Result: → MET, despite zero keyword overlap with the job ad

Every match shows the CV passage and the reasoning chain behind it.

Where It's Used

Any workflow where people are matched against requirements

The system adapts to your role profiles, not the other way around.

📥

High-volume screening

Hundreds of applications per role, every one read fully and scored consistently. The candidate at number 312 gets the same attention as number 1.

🤝

Agency & RPO recruitment

Rank candidates against each client's profile and hand over shortlists with documented, evidence-backed reasoning per candidate.

🗂️

Talent-pool mining

Re-score past applicants against new vacancies. The strong candidate who was second choice last year surfaces automatically for the next role.

🔄

Internal mobility

Match internal profiles against open roles and project staffing needs, consistently, and with the same evidence trail as external hiring.

📬

Tender & bid staffing

Verify proposed consultant CVs against tender requirements (certifications, years of experience, language levels) before submission.

🎓

Graduate & volume intakes

Score thousands of early-career applications against potential-based criteria, with documented reasoning replacing arbitrary cut-offs.

Why It Matters

Real impact on recruiting productivity and quality

5×
Faster than manual screening

What used to take a recruiter a full week of CV reading now takes under an hour, with every applicant read fully, not skimmed.

100%
Evidence-traced scores

Every score cites the exact CV passage that produced it. No black-box rankings: every shortlist decision is explainable and defensible.

∞
Role profiles per deployment

Define new requirements in plain English for each vacancy. No schema migrations, no retraining, no vendor involvement to add a role.

Common Questions

Straight answers

Is this compliant with the EU AI Act and GDPR?+
CV ranking is a high-risk use case under the EU AI Act, and the system is designed for exactly that reality: it is decision support, not automated decision-making. Every score is evidence-traced and explainable, a human reviews and can override every output, all runs are logged, and rejected candidates are never auto-declined by the system. Everything runs inside your Azure tenant under your GDPR controls, and we support your documentation obligations (risk assessment, logging, human-oversight design) as part of deployment.
How does it avoid bias in scoring?+
Scores are based only on the explicit criteria you define, and you can see exactly which CV evidence drove each score. Attributes like name, photo, age, or gender are not part of any criterion, and the visible reasoning chain means any questionable pattern can be detected and corrected, which is impossible with black-box ranking. The system is also conservative: where evidence is ambiguous, it flags "Partial" with an explanation rather than guessing.
Our requirements change with every vacancy. Does that require reconfiguration?+
No. Role requirements are defined in plain language for each vacancy: must-haves, nice-to-haves, and weights. Opening a new role is as simple as writing a new profile. No code changes, no vendor involvement, no retraining. This is a core design principle of the system.
Candidate data is highly sensitive. How is it protected?+
Everything runs within your Azure tenant. CVs are processed by your Azure AI Foundry deployment, and no candidate data is sent to external APIs. Your existing Azure AD access controls, retention policies, and data residency settings apply from day one. We provide a Data Processing Agreement for your procurement and legal review.
We already have an ATS. Why do we need this?+
Your ATS manages the pipeline; it doesn't read CVs. Its keyword filters reject candidates who used the wrong words and pass through CVs stuffed with the right ones. This system does the reading and reasoning your recruiters currently do manually, and feeds a ranked, evidence-backed shortlist into the workflow your ATS already manages. It complements your ATS, it doesn't replace it.
How long does a typical deployment take?+
A pilot on a live vacancy with your real applicant pool typically takes 2–3 weeks from kickoff. This gives you a working system on your actual CVs, with your actual requirements, not a demo. Full production deployment, if you choose to proceed, typically follows within 4–6 additional weeks depending on ATS integration requirements.
Ready to see it working?

Run it on your applicants.
Not on a slide deck.

The best way to evaluate this system is a live demo on a real, even anonymised, applicant pool from one of your actual vacancies. Book a 30-minute call and we'll show you a live ranking, then tell you honestly whether your use case is a fit.

Book a Demo Call → Download Solution Brief

No commitment · No slide deck · Engineers talking through your problem · Based in Switzerland, operating across EU & UK