CareerDEX

AI-powered career intelligence platform — job matching, resume analysis, salary prediction, and career path recommendations.

Built on dataenginex and datadex.

---

Quick Start

git clone https://github.com/TheDataEngineX/careerdex && cd careerdex
uv sync
uv run poe dev                       # Start API → http://localhost:8003

CLI

careerdex serve --port 8003          # Start CareerDEX API server
careerdex ingest --source linkedin   # Run job ingestion pipeline
careerdex match resume.pdf           # Match resume against job database
careerdex predict-salary resume.pdf  # Predict salary range
careerdex career-path "Data Analyst" # Show career progression paths

Features

Feature	Description
Multi-source ingestion	Pull jobs from LinkedIn, Indeed, Glassdoor + local JSON/CSV connectors
Resume intelligence	NLP-based skill extraction and experience normalization
Semantic matching	Embedding-based cosine similarity + skill overlap scoring
Salary prediction	XGBoost models with percentile ranges and confidence scores
Skill gap analysis	TF-IDF based recommendations for target roles
Career paths	Graph-based career trajectory recommendations
Churn prediction	Logistic regression for user engagement risk scoring
Real-time alerts	Slack notifications for pipeline events

API Endpoints

Endpoint	Method	Description
/health	GET	Health check
/ready	GET	Readiness probe
/metrics	GET	Prometheus metrics
/api/v1/careerdex/salary/prediction	POST	Salary prediction
/api/v1/careerdex/insights/skill-gaps	GET	Skill gap analysis
/api/v1/careerdex/market/careers	GET	Career path recommendations
/api/v1/careerdex/insights/career-health	GET	Career health score
/api/v1/careerdex/market/trends	GET	Market trends
/api/v1/careerdex/jobs/recommendations	GET	Job recommendations
/api/v1/models	GET	List ML models
/api/v1/predict	POST	Run ML prediction

Architecture

LinkedIn / Indeed / Glassdoor APIs
Local JSON / CSV Files
         ↓
  Job Ingestion Pipeline (Phase 2)
  Connectors → Normalization → Dedup → Quality Scoring
         ↓
  Bronze (raw) → Silver (cleaned) → Gold (enriched)
         ↓
  ML Models (Phase 4)
  ┌──────────────────────────────────────┐
  │  Resume-Job Matcher · Salary Pred.   │
  │  Skill Gaps · Career Paths · Churn   │
  └──────────────────────────────────────┘
         ↓
  FastAPI Service (:8003)

Project Structure

careerdex/
├── src/careerdex/
│   ├── api/                 # FastAPI app + routers
│   │   ├── main.py          # App entry point with middleware stack
│   │   └── routers/         # ML serving + v1 data endpoints
│   ├── core/                # Schemas, validators, settings, exceptions
│   ├── config/              # Pipeline configuration (job_config.json)
│   ├── dags/                # Airflow DAGs
│   ├── phases/              # Phased implementation (1-6)
│   ├── cli.py               # Click CLI
│   └── plugin.py            # DataEngineX plugin entry point
├── tests/
│   ├── unit/                # Unit tests
│   ├── integration/         # Integration tests (live server)
│   └── fixtures/            # Test data files
├── docs/                    # MkDocs documentation
├── .github/workflows/       # CI/CD pipelines
├── pyproject.toml           # Project configuration
├── poe_tasks.toml           # Task runner (25+ tasks)
├── Dockerfile               # Multi-stage container (port 8003)
└── docker-compose.yml       # Full stack with observability

Development

git clone https://github.com/TheDataEngineX/careerdex && cd careerdex
uv sync

# Quality checks
uv run poe lint              # Ruff lint
uv run poe typecheck         # mypy --strict
uv run poe check-all         # lint + typecheck + test

# Testing
uv run poe test              # All tests
uv run poe test-cov          # With coverage

# Run
uv run poe dev               # Dev server (port 8003)
uv run poe docker-up         # Full stack (API + Prometheus + Grafana + Jaeger)

Docker

uv run poe docker-build      # Build image
uv run poe docker-up         # Start full stack
uv run poe docker-down       # Stop everything

Services: CareerDEX API (:8003), Prometheus (:9090), Grafana (:3000), Jaeger (:16686)

---

Part of TheDataEngineX ecosystem | License: MIT