An open, JSON-based protocol for representing job applicant profiles in a structured, interoperable, and extensible manner.

🌐 Website: app-protocol.org | 📦 npm: applicant-profile-protocol

There are many competing applicant profile formats — JSON Resume, Europass, HR-XML, and countless proprietary ATS schemas. Each has limitations:

• JSON Resume — Developer-friendly but limited scope. No confidence metrics or verification.
• Europass — EU-focused, not globally adopted. Tied to government infrastructure.
• HR-XML — Enterprise-grade but heavyweight XML. Complex to implement.
• ATS Vendors — Proprietary, fragmented. No interoperability between systems.

APP unifies these into a single JSON protocol that is:

• ✅ Interoperable — Export to all major formats
• ✅ Machine-readable — JSON Schema validation, AI-ready
• ✅ Extensible — Layered architecture (core + enrichment + evidence)
• ✅ Open — Community-governed, Apache 2.0 license

# Install globally for CLI usage
npm install -g applicant-profile-protocol

# Or install locally in your project
npm install applicant-profile-protocol

# Configure npm to use GitHub Packages for @caglarorhan scope
npm config set @caglarorhan:registry https://npm.pkg.github.com

# Install globally
npm install -g @caglarorhan/applicant-profile-protocol

# Or locally
npm install @caglarorhan/applicant-profile-protocol

# Validate an APP profile
app validate your-profile.json

# Export to JSON Resume format
app export:jsonresume your-profile.json > resume.json

# Export to Europass XML
app export:europass your-profile.json > europass.xml

# Export to HR-XML
app export:hrxml your-profile.json > hrxml.xml

# Export to JSON-LD (Schema.org)
app export:jsonld your-profile.json > profile.jsonld

This repository contains the specification, JSON Schema, validation tools, and examples for the Applicant Profile Protocol (APP).

• SPEC.md: Full protocol specification
• schema/app.schema.json: JSON Schema (v1.0) for validating APP profiles
• CHANGELOG.md: Version history and release notes

• examples/minimal.json: Minimal valid APP profile
• examples/full.json: Full featured example with semantic and evidence layers
• docs/EXAMPLES.md: 10 detailed real-world examples across industries (Software Engineer, Nurse, Marketing Manager, Data Scientist, Mechanical Engineer, UX Designer, Financial Analyst, Sales Rep, Teacher, Recent Graduate)

• docs/mappings/: Export mapping guides for JSON Resume, Europass, HR-XML
• docs/semantic/json-ld.md: Guidance for the optional JSON-LD semantic layer
• src/: Validator and exporters (Node.js)
• public/tools/: Online tools for creating, validating, and converting profiles

You can validate APP JSON using common validators.

npm init -y
npm install ajv@8 ajv-formats
node -e "const Ajv=require('ajv');const addFormats=require('ajv-formats');const fs=require('fs');const schema=JSON.parse(fs.readFileSync('schema/app.schema.json'));const data=JSON.parse(fs.readFileSync('examples/full.json'));const ajv=new Ajv({allErrors:true});addFormats(ajv);const validate=ajv.compile(schema);console.log(validate(data)?'valid':'invalid',validate.errors||'');"

pip install jsonschema
python - << 'PY'
import json
from jsonschema import validate, Draft202012Validator
schema = json.load(open('schema/app.schema.json'))
instance = json.load(open('examples/full.json'))
Draft202012Validator.check_schema(schema)
validate(instance=instance, schema=schema)
print('valid')
PY

Use the browser-based tools at app-protocol.org — no installation required:

• Profile Builder — Create APP profiles using an intuitive form interface. Saves drafts to localStorage.
• Validator — Validate your APP JSON against the official schema with detailed error messages.
• Converter — Export APP profiles to JSON Resume, Europass XML, HR-XML, or JSON-LD.
• Importer — Import existing profiles from JSON Resume, Europass, HR-XML, or LinkedIn formats.
• PDF Extractor — NEW! AI-powered PDF resume extraction with OpenAI GPT-4 and Google Vision OCR. Achieves 90-95% accuracy.

We've added a powerful backend service that extracts resume data from PDF files using AI:

• AI Entity Extraction — OpenAI GPT-4 intelligently extracts contact info, experience, education, skills, projects, certifications, and more
• OCR Support — Google Cloud Vision automatically detects and processes scanned PDFs
• High Accuracy — 90-95% extraction accuracy vs 60-75% with pattern matching
• Async Processing — Bull queue with Redis handles large files efficiently
• Secure Storage — Firebase Storage + Firestore for authenticated users
• Real-time Status — Poll API for processing progress with detailed stages

1. Set up the backend (see backend/QUICKSTART.md)
2. Configure Firebase, OpenAI, and Google Cloud credentials
3. Start server: cd backend && npm run dev
4. Open public/tools/pdf-extractor-backend.html in your browser
5. Sign in and upload a PDF resume

PDF Upload → Text Extraction → OCR (if needed) → AI Extraction (GPT-4) 
  → APP Mapping → Schema Validation → Confidence Scoring → Firestore

Processing Time: 5-30 seconds depending on PDF complexity

Cost: ~$0.01 per resume (OpenAI) + ~$0.0015 per page (Vision API)

• Backend Setup Guide — Detailed configuration
• Quick Start — Get running in 5 minutes
• API Documentation — Full API reference
• Docker Setup — Container deployment

Install dependencies and run the CLI:

npm install
npx app validate examples/full.json
npx app export:jsonresume examples/full.json > dist/full.jsonresume.json
npx app export:europass examples/full.json > dist/full.europass.xml
npx app export:hrxml examples/full.json > dist/full.hrxml.xml

Or via npm scripts:

npm run validate
npm run export:jsonresume
npm run export:europass
npm run export:hrxml
npm run export:jsonld

Export APP profiles to other formats:

• ✅ JSON Resume (implemented)
• ✅ Europass XML (implemented)
• ✅ HR-XML (implemented)
• ✅ JSON-LD / Schema.org (implemented)
• 🔜 ATS-specific payloads (planned)

• Protocol uses Semantic Versioning.
• See SPEC.md for versioning policy.

Apache 2.0 — Open source, free to use and modify.

Contributions are welcome! Please open an issue or submit a pull request.

Open specification, community proposals, and backward compatibility encouraged.