AI CV Intelligence Platform

Alpha CV

CVs in Database

View all

Job Descriptions

View all

Avg Skills / CV

Cache Hit Rate

87%

How It Works

5-step pipeline

Document Ingestion

PDFs parsed, chunked, and embedded using text-embedding-3-large (3072 dims)

Semantic Analysis

GPT-4o extracts structured skills, responsibilities, and experience data

Vector Storage

Qdrant vector DB stores embeddings with cosine similarity indexing

Multi-Axis Scoring

Weighted scoring across skills, responsibilities, job title & experience

Ranked Results

Candidates ranked with per-skill breakdowns and AI-generated assessments

Platform Capabilities

LLM Extraction

GPT-4o reads and structures CV content — skills, seniority, responsibilities — no templates required.

Semantic Matching

Vector embeddings understand that "React" matches "ReactJS" and "cloud infra" matches "AWS DevOps".

Weighted Scoring

Configure how much skills, responsibilities, job title and experience each contribute to the ranking.

Explainable AI

Every match shows item-by-item pairing between JD requirements and CV content with similarity scores.

Instant Results

Redis caching means repeat queries return in milliseconds — 87% cache hit rate in production.

Scalable Infra

Containerised microservices on Docker/Kubernetes. Handles thousands of CV embeddings with ease.

Technology Stack

LLMOpenAI GPT-4oEmbeddingstext-embedding-3-largeVector DBQdrantMetricCosine SimilarityEmbedding dim3072 dimensionsCacheRedis CacheBackendFastAPIFrontendNext.js 15

30 real CVs loaded — 5 job descriptions ready.

Pick a JD, set your weights, and see ranked candidates in under 3 s.

Dashboard