Skip to content
All projects
MVPAI & RAG

AI Job Matching Engine

Job recommendation and candidate matching that combines parsed profile and job features, semantic embeddings and a 100-point scoring framework.

Overview

The matching engine behind the NextJobz / NextJobApps job portal work. It parses resumes and jobs, derives features, uses embeddings and vector search for semantic matching, and ranks results with an explicit scoring model.

Problem

Keyword matching misses candidates whose skills are phrased differently from a job description, and opaque rankings are hard to trust or tune.

Solution

Extract structured features from profiles and jobs, embed skills and text for semantic similarity, retrieve candidates or jobs by vector search, and rank them with a transparent 100-point recommendation framework.

Architecture

AI Job Matching Engine architecture, top to bottom
  1. 01

    Parsing

    • Resume parsing
    • Job parsing
  2. 02

    Features

    • Profile features
    • Job features
  3. 03

    Embedding

    • Skill / text embeddings
  4. 04

    Retrieval

    • Vector database
    • Semantic search
  5. 05

    Ranking

    • 100-point scoring framework
  6. 06

    Delivery

    • Backend APIs
    • Job portal (Next.js)
  • A 100-point scoring framework across relevant experience, duration, industry, seniority, latest category, job title, skills and education.
  • Embedding options include OpenAI text-embedding-3-small (1536 dimensions) and Qwen3-Embedding-8B (1024-dimension configurations explored), plus Sentence Transformers.
  • Semantic retrieval narrows the candidate set; the scoring layer produces the explainable ranking.

Key features

  • Resume and job parsing
  • Semantic skill and text matching
  • Explainable 100-point recommendation scoring
  • Candidate-to-job and job-to-candidate matching

Engineering challenges

  • Weighing very different signals (skills, seniority, industry, recency) into one score.
  • Matching skills semantically without over-matching loosely related terms.
  • Keeping the ranking explainable enough to tune.

Results

  • Defined a 100-point recommendation scoring framework across eight factors: relevant experience, duration, industry, seniority, latest category, job title, skills and education.