Skip to content

Capabilities

Jump to bottom
Chad Boyd edited this page Oct 18, 2025 · 1 revision

JobSentinel Current Capabilities

Version: 0.9.0
Last Updated: October 17, 2025
Status: Production Ready

This page tracks all current capabilities and features of JobSentinel. For planned features, see AI/ML Roadmap and Features Documentation.

Table of Contents

  1. Core Job Search Capabilities
  2. AI & Machine Learning
  3. Privacy & Security
  4. User Interfaces
  5. Platform Support
  6. Data Management
  7. Notifications & Alerts
  8. Developer Tools
  9. Deployment Options
  10. Accessibility

Core Job Search Capabilities

Job Sources & Scraping

JobSentinel can scrape jobs from multiple sources simultaneously:

Source Status Type Notes
Greenhouse ATS ✅ Production HTML Scraper All Greenhouse-powered company career pages
Lever ATS ✅ Production HTML Scraper All Lever-powered company career pages
JobsWithGPT ✅ Production MCP Integration 500K+ jobs via Model Context Protocol
Reed.co.uk ✅ Production API UK jobs (API key required)
JobSpy ✅ Production Multi-Aggregator Indeed, ZipRecruiter, Glassdoor, Google
LinkedIn 🧪 Beta Scraper Public job postings (no auth)
AngelList 🧪 Beta Scraper Startup jobs
RemoteOK 🧪 Beta Scraper Remote-first positions
We Work Remotely 🧪 Beta Scraper Remote job board
Hacker News 🧪 Beta Scraper Monthly "Who's Hiring" threads
Company Career Pages 🧪 Beta Generic Scraper Direct company websites

Scraping Features:

  • ✅ Concurrent scraping with rate limiting
  • ✅ Robots.txt compliance
  • ✅ Exponential backoff with jitter
  • ✅ Circuit breakers for resilience
  • ✅ Headless browser support (Playwright)
  • ✅ JavaScript-heavy site support
  • ✅ Automatic job deduplication
  • ✅ Respects API rate limits

Job Matching & Scoring

Multi-Factor Scoring Algorithm:

  • Skills Match (40%) - TF-IDF keyword matching against job description
  • Salary Range (25%) - Meets minimum salary requirements
  • Location (20%) - Remote, hybrid, on-site, city, state, country matching
  • Company Preference (10%) - Avoids denied companies
  • Recency (5%) - Prefers recently posted jobs

Matching Capabilities:

  • ✅ Configurable keyword matching
  • ✅ Keyword boost (increase weight for specific terms)
  • ✅ Keyword exclusion (filter out unwanted terms)
  • ✅ Title allowlist/blocklist
  • ✅ Location filtering (multiple formats)
  • ✅ Salary range validation with currency support
  • ✅ Company denylist
  • ✅ Customizable scoring weights
  • ✅ Score threshold alerts (default: 80+)
  • ✅ TF-IDF statistical importance

Accuracy:

  • Job matching: 87% (target: 85%)
  • Resume matching: 85% (target: 85%)

AI & Machine Learning

Natural Language Processing

BERT-Based Semantic Understanding:

  • ✅ 768-dimensional sentence embeddings
  • ✅ Sentence-BERT for similarity matching
  • ✅ 85-90% accuracy, <200ms latency
  • ✅ Automatic fallback to TF-IDF

spaCy NLP Pipeline:

  • ✅ Named Entity Recognition (NER)
  • ✅ Part-of-speech tagging
  • ✅ Skill extraction from job descriptions
  • ✅ Resume parsing and analysis

Sentiment Analysis:

  • ✅ VADER sentiment on job descriptions
  • ✅ 90%+ accuracy
  • ✅ Company culture insights

Resume Analysis

13 Industry-Specific Profiles:

  1. Technology & Software Engineering
  2. Healthcare & Medical
  3. Finance & Banking
  4. Legal & Compliance
  5. Education & Academia
  6. Sales & Business Development
  7. Marketing & Communications
  8. Engineering & Manufacturing
  9. Creative & Design
  10. Customer Service & Support
  11. Operations & Project Management
  12. Human Resources & Recruiting
  13. General/Universal

ATS Optimization Scoring (6 Dimensions):

  1. Content Quality - Impact statements, quantification
  2. Quantification - Numbers, percentages, metrics
  3. Action Verbs - Strong, specific verbs
  4. Keywords - Industry and role-specific terms
  5. Format - ATS-friendly structure
  6. Length - Optimal resume length (1-2 pages)

Resume Processing:

  • ✅ PDF and DOCX support
  • ✅ Skills extraction and categorization
  • ✅ Skills gap analysis vs job requirements
  • ✅ Learning path recommendations (Coursera, Udemy, freeCodeCamp)
  • ✅ Education and experience parsing
  • ✅ Contact information extraction
  • ✅ Resume-job matching score (85%+ correlation)

Scam & Ghost Job Detection

Pattern Detection:

  • ✅ FBI IC3 scam patterns
  • ✅ FTC fraud indicators
  • ✅ BBB scam database patterns
  • ✅ MLM/pyramid scheme detection
  • ✅ Ghost job indicators (evergreen postings, missing IDs)
  • ✅ Red flag language analysis
  • ✅ Excessive call-to-action detection

Accuracy:

  • Scam detection: 95% (target: 90%)
  • Multi-factor confidence scoring
  • Detailed explanations for flagged jobs

Optional GPT Integration

Available Models:

  • GPT-4 (~$0.03/1K tokens)
  • GPT-3.5 (~$0.0015/1K tokens)
  • Local LLaMA (7B params, <5GB RAM, free)

Capabilities:

  • 📝 Cover letter generation
  • 🤔 Interview preparation
  • 📊 Job description analysis
  • 🔄 Skills translation

Safety Features:

  • ✅ Monthly budget caps
  • ✅ Per-query limits
  • ✅ Usage tracking
  • ✅ Automatic cascade: Local LLaMA → GPT-3.5 → GPT-4

Privacy & Security

Privacy Features

  • ✅ 100% Local-First - All data stored on your machine
  • ✅ Zero Telemetry - No tracking, no analytics, no data collection
  • ✅ Privacy Dashboard - Complete data transparency and control
  • ✅ No Third-Party Sharing - Your data never leaves your machine
  • ✅ PII Redaction - GDPR/CCPA-aligned data handling
  • ✅ Local Database - SQLite stored in your user directory

Security Features

OWASP ASVS Level 2 Compliance (17 Controls):

  • ✅ Input validation on all external data
  • ✅ SQL injection prevention
  • ✅ XSS protection
  • ✅ CSRF protection
  • ✅ Rate limiting
  • ✅ Secure sessions (256-bit IDs)
  • ✅ Auto session timeout
  • ✅ Content Security Policy (CSP) with nonces

Additional Security:

  • ✅ Secrets in .env only (never committed)
  • ✅ Audit logging with HMAC-SHA256 tamper detection
  • ✅ Read-only scrapers (no writes to job boards)
  • ✅ Security scanning (Bandit, PyGuard)
  • ✅ Dependency security (Dependabot, automated scans)
  • ✅ Multi-stage Docker builds with security scanning
  • ✅ Regular security updates via auto-update

Compliance:

  • OWASP ASVS Level 2
  • GDPR data handling
  • CCPA privacy standards
  • NIST AI Risk Management Framework

User Interfaces

1. Command Line Interface (CLI)

Commands:

# Single run
python -m jsa.cli run-once

# Scheduled daemon (every 2 hours)
python -m jsa.cli run-daemon --interval 7200

# Health check
python -m jsa.cli health

# Config validation
python -m jsa.cli config-validate

# Web UI launcher
python -m jsa.cli web --port 8000

Features:

  • ✅ Full-featured command line control
  • ✅ Detailed logging and progress
  • ✅ Configuration validation
  • ✅ Health diagnostics
  • ✅ Cross-platform (Windows/macOS/Linux)

2. React Web UI (Modern)

Built With:

  • React 19
  • Vite 7
  • Tailwind CSS 4
  • Recharts (data visualization)

Features:

  • ✅ Dashboard - Job stats, recent matches, activity charts
  • ✅ Job Search - Advanced filtering, sorting, pagination
  • ✅ Application Tracker - Kanban-style management (Applied, Interview, Offer, Rejected)
  • ✅ Resume Analyzer - Upload and analyze resumes
  • ✅ Settings - Web-based configuration editor
  • ✅ Real-Time Updates - WebSocket for live notifications
  • ✅ Dark Mode - System-aware dark/light themes
  • ✅ Responsive Design - Mobile-first, all screen sizes
  • ✅ Data Visualization - Interactive charts and graphs

Accessibility:

  • ✅ WCAG 2.1 Level AA compliant
  • ✅ Full keyboard navigation
  • ✅ Screen reader support (NVDA, JAWS, VoiceOver)
  • ✅ ARIA labels and semantic HTML
  • ✅ 4.5:1 color contrast minimum
  • ✅ Clear language (8th grade reading level)

3. Flask Web UI (Legacy)

  • ✅ Alternative web interface
  • ✅ Traditional server-side rendering
  • ✅ Full feature parity with modern UI

4. GUI Launcher (Windows/macOS)

Windows:

  • ✅ Zero command line required
  • ✅ GUI buttons for all operations
  • ✅ Desktop shortcuts
  • ✅ Works without admin rights
  • ✅ Visual log viewer

macOS:

  • ✅ Native macOS integration
  • ✅ Spotlight-searchable shortcuts
  • ✅ Dock icon support
  • ✅ 5-minute setup wizard

Platform Support

Operating Systems

Platform Version Status Installation Time
Windows 11+ ✅ Production 5 minutes
macOS 15+ (Sequoia) ✅ Production 5 minutes
Linux Ubuntu 22.04+ ✅ Production 10 minutes

Installation Methods

Windows:

  • ✅ Double-click installer (launch-gui.bat)
  • ✅ Automatic dependency installation
  • ✅ No admin rights required
  • ✅ Desktop shortcuts created automatically

macOS:

  • ✅ Double-click setup wizard (setup-macos.sh)
  • ✅ Guided 5-minute installation
  • ✅ Automatic Python/Git installation
  • ✅ Spotlight-searchable shortcuts

Linux:

  • ✅ Clone and install script
  • ✅ Standard Python venv workflow
  • ✅ Package manager integration

Docker:

  • ✅ Multi-stage builds
  • ✅ Security scanning
  • ✅ Cross-platform images
  • ✅ Development and production configs

Auto-Update System

  • ✅ Zero-admin updates (Windows)
  • ✅ Automatic backup before update
  • ✅ SHA-256 verification
  • ✅ Rollback on failure
  • ✅ User notification system
  • ✅ No interruption to running jobs

Data Management

Database

SQLite (Default):

  • ✅ Zero setup required
  • ✅ Single-file database
  • ✅ Tested with 10,000+ jobs
  • ✅ ~50MB per 1,000 jobs
  • ✅ <100MB typical usage

PostgreSQL (Optional):

  • 🧪 Beta support
  • ✅ Multi-user capability
  • ✅ Cloud deployment ready
  • ✅ Advanced querying

Data Operations

Backup & Restore:

  • ✅ One-click backup
  • ✅ SHA-256 checksums
  • ✅ Automatic backup before updates
  • ✅ Point-in-time restore
  • ✅ Backup verification

Export Formats:

  • ✅ JSON (structured data)
  • ✅ CSV (spreadsheet-compatible)
  • ✅ tar.gz (compressed archives)

Database Optimization:

  • ✅ VACUUM (reclaim space)
  • ✅ ANALYZE (update statistics)
  • ✅ Index maintenance
  • ✅ Automatic optimization scheduling

Data Tracking:

  • ✅ Job persistence with full metadata
  • ✅ Score history tracking
  • ✅ Application status tracking
  • ✅ Search history
  • ✅ User activity logs

Notifications & Alerts

Supported Channels

Slack:

  • ✅ Webhook integration
  • ✅ Rich job details (title, company, location, salary, score, link)
  • ✅ Rate limiting (respects Slack API limits)
  • ✅ Retry logic with exponential backoff
  • ✅ Digest mode (batch multiple jobs)

Email:

  • ✅ SMTP support (Gmail, Outlook, custom)
  • ✅ HTML formatted emails
  • ✅ Embedded job details
  • ✅ Batch digest mode
  • ✅ Configurable frequency

In-App:

  • ✅ WebSocket real-time notifications
  • ✅ Browser notifications
  • ✅ Visual badges and counters

Alert Configuration

  • ✅ Threshold-based triggers (default: 80+ score)
  • ✅ Custom score thresholds
  • ✅ Frequency controls (immediate, hourly, daily)
  • ✅ Quiet hours support
  • ✅ Channel-specific rules

Developer Tools

Testing

Test Suite:

  • ✅ pytest framework
  • ✅ 87% code coverage (target: 85%)
  • ✅ Property-based testing (Hypothesis)
  • ✅ Mutation testing (mutmut)
  • ✅ Integration tests
  • ✅ Unit tests
  • ✅ End-to-end tests

Quality Tools:

  • ✅ Type checking (mypy strict)
  • ✅ Linting (Ruff: E, F, B, I, UP, S)
  • ✅ Formatting (Black, line-length=100)
  • ✅ Pre-commit hooks
  • ✅ MegaLinter (cross-file-type linting)

CI/CD

GitHub Actions Workflows:

  • ✅ Continuous integration
  • ✅ Code coverage reporting
  • ✅ CodeQL security scanning
  • ✅ Dependency scanning
  • ✅ Documentation linting
  • ✅ Windows deployment testing
  • ✅ macOS deployment testing
  • ✅ Path-based filtering

Performance Monitoring:

  • ✅ Test execution tracking
  • ✅ Coverage trends
  • ✅ Security scan results
  • ✅ Scorecard badges

Documentation

Comprehensive Guides (13 Essential Documents):

  1. README.md - Project overview
  2. QUICKSTART.md - Universal setup guide
  3. ARCHITECTURE.md - System design
  4. TROUBLESHOOTING.md - Common issues
  5. UI.md - User interface guide
  6. FEATURES.md - Complete feature catalog
  7. AI_ML_ROADMAP.md - AI/ML vision
  8. API_INTEGRATION_GUIDE.md - Add new job boards
  9. DATABASE_GUIDE.md - Database management
  10. DEPLOYMENT_GUIDE.md - Cloud deployment
  11. BEST_PRACTICES.md - Coding standards
  12. AUTHORITATIVE_STANDARDS.md - 45+ standards
  13. GLOSSARY.md - Terminology reference

Developer Resources:

  • ✅ API documentation
  • ✅ Plugin system guide
  • ✅ Custom scraper template
  • ✅ Contributing guidelines
  • ✅ Code of conduct
  • ✅ Security policy

Integration & Extensibility

Model Context Protocol (MCP):

  • ✅ GitHub Copilot integration
  • ✅ MCP server support
  • ✅ AI-assisted development

Plugin Architecture:

  • ✅ Modular job source plugins
  • ✅ Custom scraper support
  • ✅ Plugin template and examples

APIs:

  • ✅ RESTful API (FastAPI)
  • ✅ WebSocket API for real-time updates
  • ✅ GraphQL (planned)

Deployment Options

Local Deployment

Cost: $0
Setup Time: 5-10 minutes

  • ✅ Run on your personal machine
  • ✅ Complete data privacy
  • ✅ No ongoing costs
  • ✅ GUI launcher (Windows/macOS)
  • ✅ CLI for advanced users

Docker

Cost: $0 (local)
Setup Time: 5 minutes

  • ✅ Multi-stage builds
  • ✅ Security scanning integrated
  • ✅ Development and production configs
  • ✅ Cross-platform images
  • ✅ Docker Compose support

Cloud Platforms

AWS Lambda:

  • ✅ Serverless deployment
  • ✅ EventBridge scheduling
  • ✅ Terraform templates
  • Cost: ~$5-10/month

GCP Cloud Run:

  • ✅ Managed containers
  • ✅ Auto-scaling
  • ✅ Terraform templates
  • Cost: ~$8-15/month

Azure Container Instances:

  • ✅ Serverless containers
  • ✅ Fast deployment
  • ✅ Terraform templates
  • Cost: ~$10-20/month

Kubernetes:

  • ✅ Helm charts included
  • ✅ Enterprise-ready
  • ✅ Multi-tenant support
  • Cost: ~$50+/month (cluster costs)

Infrastructure as Code

  • ✅ Terraform templates for all platforms
  • ✅ Environment-specific configurations
  • ✅ Automated deployment pipelines
  • ✅ Infrastructure documentation

Accessibility

WCAG 2.1 Level AA Compliance

Visual Accessibility:

  • ✅ 4.5:1 minimum color contrast
  • ✅ Resizable text (up to 200%)
  • ✅ No information conveyed by color alone
  • ✅ High contrast dark mode

Keyboard Accessibility:

  • ✅ Full keyboard navigation
  • ✅ Visible focus indicators
  • ✅ Logical tab order
  • ✅ Keyboard shortcuts

Screen Reader Support:

  • ✅ NVDA (Windows)
  • ✅ JAWS (Windows)
  • ✅ VoiceOver (macOS/iOS)
  • ✅ ARIA labels on all interactive elements
  • ✅ Semantic HTML structure

Content Accessibility:

  • ✅ Clear, concise language (8th grade level)
  • ✅ Descriptive link text
  • ✅ Alternative text for images
  • ✅ Proper heading hierarchy

Additional Features:

  • ✅ Focus management for dynamic content
  • ✅ Skip navigation links
  • ✅ Form labels and error messages
  • ✅ Status messages for screen readers

Performance Metrics

Current Performance (v0.9.0)

Metric Target Actual Status
Job Matching Accuracy 85% 87% ✅ Exceeds
Scam Detection Accuracy 90% 95% ✅ Exceeds
Resume Analysis Accuracy 85% 85% ✅ Meets
Semantic Matching Latency <200ms <200ms ✅ Meets
Test Coverage 85% 87% ✅ Exceeds
Type Safety 100% 100% ✅ Meets
Security Issues 0 0 ✅ Meets

Resource Usage

  • Scrape Time: 30-60s per source
  • Scoring: <5s per 100 jobs
  • Alert Latency: <10s (Slack/email)
  • Memory: <500MB (includes ML models)
  • Disk: ~50MB per 1,000 jobs
  • Database: <100MB typical

Configuration System

Configuration File Format

Location: deploy/common/config/user_prefs.json
Format: JSON with schema validation

Available Settings

Job Matching:

  • keywords - Terms to match in title/description
  • locations - Allowed locations (Remote, cities, states)
  • min_salary - Minimum annual salary (USD)
  • denied_companies - Companies to exclude
  • title_allowlist - Include specific job titles
  • title_blocklist - Exclude specific job titles
  • keyword_boost - Increase weight for terms
  • keyword_exclude - Filter out terms

Job Sources:

  • Per-source enable/disable
  • API keys where required
  • Custom scraper configuration
  • Rate limiting settings

Scoring Weights:

  • Customize factor weights (skills, salary, location, etc.)
  • Set alert threshold
  • Configure matching algorithm

Alerts:

  • Slack webhook URL
  • Email SMTP configuration
  • Digest mode settings
  • Quiet hours

Configuration Management

  • ✅ JSON schema validation
  • ✅ Hot reload (no restart required)
  • ✅ Multiple profiles support
  • ✅ Config validation CLI command
  • ✅ Web-based settings editor
  • ✅ Example configurations included

Unique Capabilities

JobSentinel is the only job search tool that offers:

  1. ✅ Privacy Dashboard - Complete data transparency
  2. ✅ Auto-Update with Rollback - Zero-admin updates for Windows
  3. ✅ Zero Admin Rights - Works on locked-down corporate computers
  4. ✅ 100% Local-First - All data stays on YOUR machine
  5. ✅ Open Source - Full source code available (MIT License)
  6. ✅ $0 Cost Baseline - Free forever with optional paid upgrades
  7. ✅ Custom Scraper Support - Add any job board yourself
  8. ✅ FBI/FTC Scam Detection - Industry-leading fraud protection
  9. ✅ 13 Industry Resume Profiles - Comprehensive ATS optimization
  10. ✅ Self-Hosted or Cloud - Your choice of deployment

Technology Stack

Backend

  • Language: Python 3.12+
  • Framework: FastAPI (REST), Flask (legacy web)
  • Database: SQLite (default), PostgreSQL (optional)
  • ML/AI: BERT, spaCy, scikit-learn, Sentence-BERT, VADER
  • Scraping: Playwright, BeautifulSoup4, Requests
  • Task Queue: APScheduler
  • Testing: pytest, Hypothesis, mutmut

Frontend

  • Framework: React 19
  • Build Tool: Vite 7
  • Styling: Tailwind CSS 4
  • Charts: Recharts
  • State: React Context API
  • WebSocket: Socket.io

Infrastructure

  • Container: Docker
  • IaC: Terraform
  • Orchestration: Kubernetes (Helm)
  • CI/CD: GitHub Actions
  • Monitoring: Health check system

Security

  • Scanning: Bandit, PyGuard, CodeQL
  • Dependencies: Dependabot, Safety
  • Secrets: python-dotenv
  • Validation: Pydantic, JSON Schema

Version History

v0.9.0 (Current - October 2025)

  • ✅ React 19 + Vite 7 modern UI
  • ✅ Consolidated documentation (13 essential guides)
  • ✅ Platform installers (Windows/macOS)
  • ✅ WCAG 2.1 AA accessibility
  • ✅ Privacy dashboard
  • ✅ Auto-update system
  • ✅ Backup & restore
  • ✅ Email notifications

v0.6.0 (September 2025)

  • ✅ Core scraping and matching
  • ✅ AI/ML capabilities (BERT, spaCy)
  • ✅ Privacy and security features
  • ✅ Resume analysis (13 profiles)
  • ✅ Scam detection (FBI/FTC patterns)
  • ✅ Modern React UI
  • ✅ OWASP ASVS compliance

v0.5.0 (August 2025)

  • ✅ Initial release
  • ✅ Basic scraping (Greenhouse, Lever)
  • ✅ Multi-factor scoring
  • ✅ SQLite database
  • ✅ Slack alerts
  • ✅ CLI interface

Future Roadmap

v1.0 (Q1 2026)

Planned Features:

  • Cross-encoder reranking for improved matching
  • GPT-4 integration (optional, cost-controlled)
  • Bias detection (gender, age, salary, location)
  • Skills taxonomy integration (LinkedIn Skills Graph)
  • Database encryption at rest
  • Browser extension
  • Mobile companion app (PWA)

For detailed roadmap, see:

Getting Started

Quick Links

  • Installation: See Quick Start Guide
  • Configuration: Edit deploy/common/config/user_prefs.json
  • First Run: python -m jsa.cli run-once
  • Web UI: python -m jsa.cli web --port 8000
  • Help: python -m jsa.cli --help

Platform-Specific Setup

Windows:

  1. Download repository ZIP
  2. Extract to Desktop
  3. Double-click deploy/local/windows/launch-gui.bat

macOS:

  1. Download repository ZIP
  2. Extract to Desktop
  3. Double-click deploy/local/macos/setup-macos.sh
  4. Follow 5-minute wizard

Linux:

git clone https://github.com/cboyd0319/JobSentinel
cd JobSentinel
python3 -m venv .venv && source .venv/bin/activate
pip install -e .
playwright install chromium
cp deploy/common/config/user_prefs.example.json deploy/common/config/user_prefs.json
python -m jsa.cli run-once

Support & Community

Getting Help

Contributing

We welcome contributions! See:

Security

Report vulnerabilities responsibly:

License

JobSentinel is open source software licensed under the MIT License.

This means you can:

  • ✅ Use commercially
  • ✅ Modify the code
  • ✅ Distribute
  • ✅ Use privately

With the requirement:

  • 📋 Include the original license and copyright notice

Acknowledgments

JobSentinel is built with:

  • Python and the amazing Python community
  • React and the modern web ecosystem
  • Open source ML models (BERT, spaCy)
  • Guidance from FBI IC3, FTC, and BBB on fraud patterns
  • OWASP security standards
  • WCAG accessibility guidelines

Made with ❤️ for job seekers who value privacy

Last Updated: October 17, 2025
Version: 0.9.0
Maintainer: Chad Boyd
Repository: https://github.com/cboyd0319/JobSentinel

Clone this wiki locally