A Complete Recruitment Platform That Takes the Leap into the Future
An AI-powered CV sorting and candidate scoring platform using data science, NLP, and ethical scoring models — deployed for a leading French bank.
Industry
Human Resources
Scope
New Product
Technologies
Region
France (EN + FR)
The Challenge
It's yet another job season and you find yourself in the same tight spot. Hit that post button and your nightmare starts — only catch, it does not end when you wake up. A relatable experience for every recruiter: getting hundreds of resumes a day, sifting through each, finding suitable candidates, shortlisting them — all while making sure they meet the company's requirements.
This prompted our client to develop a powerful decision-supporting tool — a CV sorting platform using Data Science. It combines the performance of precise sorting with concrete ethical principles. Automation of job publication, CV selection, and applicant communication has drastically reduced recruiter workload and saved significant time.
AI & NLP Capabilities
NLP Resume Parsing
Multi-language NLP extracts structured data from any CV format (English & French)
ML-Powered Scoring
Completeness and qualification criteria scored via transparent decision matrix
Ethical AI / Bias-Free
Transparent, traceable scoring — selects CVs purely on talent and skill, without bias
HR Analytics Dashboard
Chart.js-powered analytics to track performance indicators and benchmarks
Challenges
Pilot Phase — Resume Parsing Accuracy
An in-house algorithm was developed for resume parsing, achieving only 60% accuracy for English and under 50% for French. Due to inadequate CV volumes and reliance on existing models for layout analysis and NLP, accurate data extraction was impossible. Resumes in completely new formats failed to be recognized. Different resume sections (work experience, education, languages) each required a separate model.
Phase 1 — Scoring Limitations
The scoring method was of basic design, missing premium resumes due to inaccuracies in qualification scoring. The Bag of Words dictionary also had to be significantly enhanced to handle the large volume of CVs received.
Phase 2 — Ontology & Transparency
Fine-tuning ontology as an evolving process alongside recruitment profiles. A key goal: transparent scoring — selecting CVs without bias, for multiple profiles simultaneously, purely based on talent and skill required by the business.
Solution
Cloudium and the client jointly built a model deployed on AWS EC2 for a leading bank in France. The platform includes:
Technology Stack
Python Django
Angular 8
AWS EC2
How AI Powers Ethical Recruitment
The platform leverages NLP, machine learning, and ontology-driven models to transform resume chaos into transparent, bias-free talent acquisition.
Multi-Language NLP Parsing
Advanced NLP engine (Textkernel) extracts structured data from any resume format in both English and French, handling varied layouts and sections intelligently.
Ontology-Driven Matching
An evolving knowledge graph (ontology) maps job profiles, skills, and qualifications — enabling nuanced matching beyond simple keyword search.
Ethical & Bias-Free AI
Transparent, fully traceable scoring parameters ensure candidates are selected purely on talent and skill — free from demographic or unconscious bias.
Decision Matrix Scoring
Resumes scored along Completeness and Qualification criteria and displayed in an actionable matrix with discriminating thresholds for recruiters.
Automated CV Workflow
End-to-end automation of job publication, CV ingestion, classification, applicant communication, and shortlist generation — drastically reducing recruiter workload.
Real-Time HR Analytics
Chart.js-powered dashboards visualize recruitment KPIs, pipeline health, scoring distributions, and time-to-hire metrics in real time.
Return on Investment
Thousands of students graduate each year worldwide. Millions of resumes reach an HR's desk every day. A few hundreds get placed — but most are not even in the right industry. This is what our client solves with their recruitment platform, and why they reached out to Cloudium.
With our ever-driven task force and top-of-the-line technical expertise, Cloudium has a track record of success stories. Here's what we delivered:
- Structured Data Extraction — from the large number of resumes using advanced NLP parsing
- Ontology-Based Scoring & Matching — on the basis of recruitment profiles derived from features contained in the Ontology
- Premium Resume Collection — curated shortlists that aptly match a company's business needs
- Right-Candidate Assurance — ensuring every selection is the right choice through transparent, traceable AI scoring
Need an AI-Powered HR Platform?
From NLP-driven resume parsing to ethical AI scoring — let Cloudium build your next intelligent recruitment solution.
Start a Conversation