STYX Guard — Enterprise AI Security

Stop employees from leaking sensitive data to ChatGPT, Claude, and Gemini. Real-time DLP for AI tools.

Your team uses ChatGPT. They use Claude. They use Gemini. And they're pasting customer data, API keys, source code, and passwords into these tools without thinking twice. STYX Guard catches it before it leaves your browser. WHAT STYX DOES When an employee types something sensitive into an AI tool, STYX intercepts it. You decide what happens next — block the message, warn the user, redact the data automatically, or log it silently. All four options are available per rule. Set rules once. They apply across every AI tool your team uses. No training, no friction, no productivity hit. WHAT GETS DETECTED Out of the box, STYX detects: - Credit card numbers, SSNs, Aadhaar, phone numbers - Passwords, API keys, JWT tokens, AWS credentials - Custom keywords and regex patterns you define - File uploads (blocks drag-and-drop, paste, and attach) WHICH AI TOOLS Today: Claude (claude.ai), ChatGPT (chatgpt.com), Gemini (gemini.google.com) Coming soon: Microsoft Copilot, DeepSeek, Perplexity Or add any custom LLM by URL — internal AI tools, private deployments, or new platforms as they launch. HOW IT'S USED Admins set up rules at app.hellm.io. The dashboard shows every violation in real time — who triggered it, on which tool, with what content. Rules can apply to your whole org, a specific team like Finance, or individual users. Employees install STYX, log in once with the credentials from their welcome email, and they're protected. They can't disable it, can't see other people's data, can't modify policies. Everything stays in the admin's control. PRIVACY STYX only runs on AI platform domains. Nothing else. We don't track browsing history, don't sell data, don't serve ads, don't share with third parties. Only 200-character snippets of flagged messages are captured. Full conversations are never recorded. The extension is built on Manifest V3 — Chrome's most secure standard. All traffic is HTTPS, passwords are bcrypt-hashed, authentication is JWT-based. WHO IT'S FOR Security teams worried about AI data leaks. IT admins rolling out AI governance. Compliance officers meeting HIPAA, GDPR, PCI-DSS, SOC2. Companies that want to enable AI productivity without giving up control. We work with financial services, healthcare, legal, government, tech, e-commerce, and consulting firms. GETTING STARTED Free to use. No credit card needed. 1. Sign up at hellm.io 2. Invite your team from the dashboard 3. Set up your rules (templates included) 4. Done Questions? support@hellm.io — we respond within a day. Website: www.hellm.io Admin Console: awww.pp.hellm.io Privacy: www.hellm.io/privacy STYX Guard — nothing passes without permission. Built by heLLM.