STING (Secure Tactical Intelligence Network Gateway) is an enterprise AI platform that lets you deploy powerful LLMs on your own infrastructure while maintaining complete data sovereignty. Unlike cloud AI services, STING runs entirely on-premises, ensuring your sensitive data never leaves your servers.

STING also introduces Honey Jars - secure knowledge containers that let you package your expertise and optionally monetize it through our marketplace (Enterprise edition).

Running an LLM locally gives you a generic AI. STING gives you YOUR AI—trained on your data, following your rules, integrated with your tools. Plus, you can package and sell your expertise. We are the first platform to offer this flexibility.

Key Differences:

• Knowledge Management: Honey Jar technology for secure knowledge packaging
• Enterprise Security: 50+ PII types protected, audit logs, compliance reporting
• Business Intelligence: Learns from your specific data and processes
• Monetization: Sell your expertise through marketplace (Enterprise)
• Production Ready: Authentication, monitoring, health checks all built-in
• Hybrid Deployment: Mix local and cloud AI based on your policies

Yes! STING Community Edition is completely free, forever. No credit card required, no time limits, no feature trials that expire.

What's included in CE:

• Full self-hosted deployment
• Basic Bee AI assistant
• Up to 3 knowledge bases (Honey Jars)
• Basic PII scrambling
• Community support

What's not included:

• Marketplace access (buy or sell knowledge)
• Advanced PII protection
• Full audit logs
• Enterprise directory integration
• Commercial support

Pro and Enterprise editions add advanced features, marketplace access, and professional support.

A Honey Jar is a secure knowledge container—think of it as an encrypted package that holds your documents, processes, and expertise together.

Key Features:

• Encrypted & Secure: AES-256 encryption protects your data
• AI-Searchable: Bee AI can find answers without exposing raw files
• Fine-Grained Access: Control exactly who can use each knowledge base
• Optionally Sellable: Monetize your expertise on the marketplace (Enterprise)

Simple example: Your legal team creates a "Contract Templates" Honey Jar containing 500 contracts. When sales asks "find NDA template for California," Bee AI searches the jar and provides the answer - without anyone seeing the raw files. The knowledge is accessible, but the data stays protected.

Why "Honey Jar"? Just like bees store nectar (knowledge) in honeycomb cells, STING stores your organizational knowledge in secure containers 🐝

STING uses multi-layered security to ensure your data stays protected:

Data Protection:

• Complete data sovereignty: Your data never leaves your infrastructure
• AES-256 encryption at rest: All stored data is encrypted
• TLS 1.3 in transit: All network communications encrypted
• End-to-end encrypted Honey Jars: Knowledge bases stay encrypted

PII Protection:

• 50+ PII types detected: Names, SSNs, credit cards, medical records, etc.
• Advanced technology: Scrambles data before AI processing
• AI report review: Final check before any data leaves (Enterprise)
• Passkey authentication: WebAuthn required for sensitive operations

Access Control:

• Zero trust architecture: Never trust, always verify
• Role-based access control (RBAC): Fine-grained permissions
• Multi-factor authentication: Required for all sensitive operations
• Complete audit trails: Every action logged and traceable

View full security framework →

STING supports a wide range of AI models, both local and cloud-based:

Local Models (Included):

• Microsoft Phi-3: Default model, GPT-3.5 level performance, great for general use
• DeepSeek Coder: State-of-the-art code generation
• TinyLlama: Lightweight, fast, runs on low-resource systems
• Any Ollama-compatible model: Llama 3, Mistral, Mixtral, etc.

Cloud AI (Optional - Pro+):

• OpenAI: GPT-4, GPT-3.5-turbo
• Anthropic: Claude 3 family
• Google: Gemini Pro
• Cohere: Command models
• Custom: Any OpenAI-compatible API endpoint

Hybrid Deployment: Use policy-based routing to control which data uses which models. Keep sensitive processing local while leveraging cloud AI for complex reports or non-sensitive tasks.

STING is designed for rapid deployment:

• Download to first query: 30 minutes
• Full production deployment with your data: 2-5 days
• Enterprise setup with security review: 1-2 weeks

One-Command Install:

Our bootstrap script handles everything automatically:

bash -c "$(curl -fsSL https://raw.githubusercontent.com/AlphaBytez/STING-CE-Public/main/bootstrap.sh)"

This command will:

• ✅ Detect your platform (macOS, WSL, Debian/Ubuntu)
• ✅ Clone the repository and check system requirements
• ✅ Launch the web-based setup wizard
• ✅ Guide you through configuration
• ✅ Install and start all services automatically

View full installation guide →

Yes, absolutely! Once installed, STING can run completely offline.

Air-Gapped Deployment:

• ✅ No phone-home, no telemetry, no external API calls
• ✅ Zero runtime dependencies on external services
• ✅ All AI processing happens locally
• ✅ Perfect for classified or secure environments
• ✅ Government and military approved architectures

Note: Initial installation requires internet to download models and dependencies. After that, fully offline operation is supported.

Hybrid Option: You can also configure STING to use cloud AI services for specific tasks while keeping sensitive data processing local. You control the policies.

Not at all! STING's hybrid architecture lets you use both.

How It Works:

• Policy-Based Routing: You configure rules for what uses which AI
• Sensitive Data Stays Local: Automatically route confidential data to local models
• Cloud for Complex Tasks: Use GPT-4 or Claude for advanced analysis when appropriate
• Automatic Fallback: If cloud is unavailable, STING falls back to local models

Example Policy:

"All HR and legal documents must use local models. Marketing content can use cloud AI for report generation."

Coming Soon: Our Cloud Pro and Cloud Enterprise editions will offer fully managed cloud deployment with isolated sandboxes for teams that prefer zero infrastructure management.

Minimum Requirements:

• OS: Ubuntu 20.04+, Debian 11+, macOS, WSL2
• RAM: 8GB (16GB+ recommended)
• CPU: 4 cores (8+ recommended)
• Storage: 50GB free (100GB+ SSD recommended)
• GPU: Optional (NVIDIA GPU for faster inference)
• Docker: Auto-installed if needed

Recommended for Production:

• RAM: 16-32GB
• CPU: 8-16 cores
• Storage: 100GB+ NVMe SSD
• GPU: NVIDIA GPU with 8GB+ VRAM (RTX 3060 or better)

Cloud Deployment: Works on AWS, Azure, GCP, or any cloud provider. We provide Terraform templates for easy deployment.

You absolutely could piece together a DIY solution with open-source tools. Many technical teams start there.

But here's what STING gives you that DIY doesn't:

• Plug-and-play setup: One command install vs weeks of integration work. Get running in 30 minutes, not 30 days.
• Production-ready from day one: Audit logs, health checks, monitoring, compliance reporting all built-in. DIY means building all this yourself.
• Built-in PII protection: 50+ types automatically detected and scrambled (not just regex patterns). Enterprise-grade data protection out of the box.
• Enterprise authentication: WebAuthn passkeys, MFA, RBAC, session management via Ory Kratos. DIY means rolling your own auth (or trusting a random npm package).
• Knowledge marketplace: Monetize your expertise by selling Honey Jars. This is unique to STING - you can't do this with DIY.
• Support when you need it: Pro/Enterprise gets you real humans who know the codebase. DIY means Stack Overflow at 2am.

Bottom line: STING is the turnkey solution. DIY is cheaper upfront but costs you engineering time, security risks, and missed revenue opportunities. If you're a hobbyist or want a learning project, DIY is great. If you're a business that needs secure AI running this month, STING saves you months of work.

Each edition builds on the previous one with additional features and support:

STING CE (Community - Free Forever):

• ✅ Self-hosted deployment
• ✅ Basic Bee AI assistant
• ✅ Up to 3 knowledge bases
• ✅ Basic PII scrambling
• ❌ No marketplace access
• ⚠️ Limited audit logs
• 📧 Community support only

STING Pro (Coming Q2 2025):

• ✅ Everything in CE, plus:
• ✅ Advanced Bee AI
• ✅ Unlimited knowledge bases
• ✅ Buy from marketplace
• ✅ Advanced PII scrambling
• ✅ Full audit logs
• ✅ Enterprise directory integration
• 📧 Email support (48h SLA)

STING Enterprise (Coming Q3 2025):

• ✅ Everything in Pro, plus:
• ✅ Sell on marketplace (85% revenue share)
• ✅ Custom AI models
• ✅ Cloud + on-prem hybrid options
• ✅ Enhanced compliance (SOC2, ISO)
• ✅ AI-powered report review
• ✅ HashiCorp Vault integration
• 📞 24/5 priority support (4h SLA)
• 👤 Dedicated success manager

View full feature comparison →

With STING Enterprise, you can monetize your expertise through our knowledge marketplace:

How It Works:

1. Package Your Expertise: Create Honey Jars from your documents, processes, and institutional knowledge
2. Set Your Price: Choose pricing model (one-time, subscription, usage-based)
3. List on Marketplace: Publish to STING's secure marketplace
4. Earn Revenue: Get 85% of all sales, monthly payouts

Revenue Potential:

• Legal Expertise: Sell specialized case law analysis ($500-5K/month typical)
• Industry Frameworks: Consulting methodologies ($1K-10K/month typical)
• Technical Knowledge: Specialized technical documentation ($100-2K/month typical)
• Compliance Templates: Industry-specific compliance knowledge ($500-3K/month typical)

You maintain complete control:

• Set your own pricing and terms
• Choose who can access (public, approved buyers only, etc.)
• Update knowledge bases anytime
• Full analytics on usage and revenue
• Your IP stays protected with encryption

*Marketplace selling available in Enterprise edition only. Expected Q3 2025.

Yes! We're committed to making STING accessible to educational institutions and non-profit organizations.

Current Status:

• ✅ STING CE is completely free for everyone, including education and non-profits
• ⏳ Special pricing for Pro/Enterprise editions planned for 2025 launch
• ⏳ Grants program for qualifying organizations in development

Coming Soon:

• Academic research licenses
• Non-profit organization discounts
• Student developer access programs
• Educational institution site licenses

Contact us about educational pricing →

Getting started with STING is simple:

Option 1: Try STING CE (Free)

1. Run the one-command installer (works on macOS, Linux, WSL)
2. Follow the web-based setup wizard
3. Upload your first documents
4. Start asking Bee AI questions

Install command:

bash -c "$(curl -fsSL https://raw.githubusercontent.com/AlphaBytez/STING-CE-Public/main/bootstrap.sh)"

View full installation guide →

Option 2: Schedule a Demo

Want to see STING in action first? Schedule a personalized demo with our team:

• See real-world use cases
• Ask technical questions
• Discuss your specific requirements
• Learn about Pro and Enterprise options

Schedule a demo →

Option 3: Join Early Access

Get priority access to Pro and Enterprise editions when they launch:

• Exclusive early access pricing
• Priority onboarding with implementation team
• Beta access to new features
• Marketplace credits

Join early access program →

Yes! Training and onboarding vary by edition:

Community Edition:

• 📚 Comprehensive documentation
• 🎥 Video tutorials (coming soon)
• 💬 Community forum support
• 📖 Self-service knowledge base

Pro Edition:

• ✅ Email onboarding guidance
• ✅ Setup assistance
• ✅ Access to training materials
• 📧 Email support (48h response)

Enterprise Edition:

• ✅ Dedicated onboarding specialist
• ✅ Custom training sessions for your team
• ✅ On-site deployment assistance (optional)
• ✅ Ongoing support and training
• 📞 24/5 priority support
• 👤 Dedicated success manager

Contact us about training options →

Support options depend on your edition:

Community Edition:

• Community forum support
• GitHub issue tracking
• Documentation and guides
• Best-effort response time

Pro Edition:

• Email support (business hours)
• 48-hour response SLA
• Knowledge base access
• Community forum priority

Enterprise Edition:

• 24/5 priority support
• 4-hour response SLA
• Dedicated success manager
• Scheduled support calls
• On-demand training sessions
• Custom integration assistance

Enterprise Swarm:

• White glove support team
• 1-hour response SLA
• On-site deployment assistance
• Custom training programs
• Executive business reviews
• 24/5 phone support hotline

Compare support options →