How AI-Powered Talent Matching Works at Pelincor

At Pelincor, we’ve built a sophisticated talent matching system specifically designed for the mining industry. Here’s how it works under the hood.

Step 1: Intelligent CV Parsing

When a candidate uploads their CV, our AI (powered by GPT-4o-mini) extracts structured data including:

Personal information and contact details
Skills with proficiency levels
Work experience with responsibilities
Education and certifications
Languages spoken

This process takes seconds and achieves 95%+ accuracy in data extraction.

Step 2: Semantic Embeddings

We convert both talent profiles and job postings into high-dimensional vectors (384 dimensions) using OpenAI’s text-embedding-3-small model. These embeddings capture the semantic meaning of a profile — not just keywords, but the deeper context of someone’s experience.

Step 3: Multi-Dimensional Matching

Our matching engine evaluates candidates across six dimensions:

Skills Match (30%): How well do the candidate’s skills align with job requirements?
Experience Match (15%): Years and relevance of experience
Education Match (5%): Educational background alignment
Location Match (5%): Geographic proximity or remote work compatibility
Salary Match (5%): Compensation expectations vs. offering
Semantic Similarity (40%): Deep contextual matching using AI embeddings

Step 4: Detailed Analysis

For each match, GPT-4o generates a comprehensive analysis including:

Overall match score (0-100%)
Breakdown by dimension
Key strengths for the role
Gaps and development areas
Recommendations for both candidate and employer

The Result

Companies using Pelincor report a 3x improvement in candidate quality and a 60% reduction in time-to-hire. Candidates receive highly relevant job recommendations with an average match accuracy of 95%.