PLATFORM FEATURES

One platform. Every AI blind spot covered.

Shadow AI discovery, sensitive data detection and policy enforcement — from browser to IDE to workspace.

Book a 30-minute callVerify a bundle yourself

297 AI providers · 9,000+ models · 22 data patterns

Discovery & Visibility

See Every AI Tool Across Your Organization

Dual-channel discovery combines agentless workspace scanning with browser-level capture. No blind spots between sanctioned and shadow AI.

Google Workspace + M365 Connectors

Agentless OAuth discovery. Connect in 2 minutes. Token audit reveals every AI app employees have authorized — zero agent install.
Learn more →

Browser Extension (Chrome + Edge)

MV3 extension captures prompts, responses, and model selection in real time. Published on Chrome Web Store and Edge Add-ons.

MCP Gateway for AI Coding Assistants

stdio proxy between AI agents and MCP servers. Supports Cursor, Claude Code, VS Code, Windsurf, Zed, JetBrains and 15 more — 21 in the gateway’s discovery registry.
Learn more →

297 AI Provider Detection

Detects ChatGPT, Claude, Gemini, Perplexity, Midjourney and 292 more — 297 in the registry. Registry updated daily from models.dev.

9,000+ Model Registry with Pricing

Every model tracked with per-token pricing, capabilities, and release dates. Cost estimation on every interaction.

Sensitive Data Detection

Catch Data Leaks Before They Leave

Real-time scanning across every capture channel. Built-in patterns for the data types that matter most, plus custom patterns for your needs.

22 Sensitive Data Patterns

SSNs, credit cards (Luhn validated), API keys, passwords, PHI, PII, source code, AWS credentials. Each pattern tuned for precision.
Learn more →

Custom Pattern Support

Add regex patterns for internal project codes, customer IDs, or proprietary formats. Runs alongside built-in detectors with the same scoring.

Real-Time Scanning

Pre-send evaluation under 150ms server-side. Extension intercepts before the request leaves the browser. MCP Gateway scans inline.

Policy Enforcement

Enforce Before the Prompt Leaves

Pre-send interception with three enforcement modes. Policies scope to departments, providers, or individual users with 9 condition operators.

Block, Redact, or Coach

Block prevents requests entirely. Redact strips sensitive data inline. Coach shows an educational overlay with edit-and-resubmit.
Learn more →

Department + Provider Scoped Rules

Target policies to departments, user emails, AI providers, or MCP server names. Engineering uses Copilot freely while Finance is restricted.

9 Condition Operators

exists, equals, not_equals, contains, not_contains, greater_than, less_than, in, not_in. Combine with AND logic for precise targeting.

Coaching Overlay with Edit Capability

Users see exactly what was detected and why. They can edit their prompt to remove sensitive data and re-submit without losing context.

Enterprise Security

Built for Security Teams, Not Just Compliance

SSO, RBAC, encryption at rest, immutable audit trails, and exportable reports. Everything your CISO will ask for.

SSO via WorkOS

Okta, Azure AD, Google Workspace, and any SAML/OIDC provider. Onboarding in 5 minutes.

RBAC with 4 Roles

Admin, Security Analyst, Compliance Officer, and Viewer with granular page-level access. Each role sees only what they need. All changes logged.

AES-256-GCM Encryption

All prompts and responses encrypted at rest. Encryption key required in production. Graceful degradation in development.

Full Audit Trail

Administrator actions, policy changes, enforcement decisions, sign-ins and connector events are logged with actor, action, resource and timestamp. Exported audit bundles support offline verification against a published JWKS — no Vloex account needed.
Learn more →

Reports (CSV + HTML)

Scheduled and on-demand. Export AI usage by department, provider, risk level. HTML for executives, CSV for analysts.

Deployment Options

Deploy Your Way. Every Channel Covered.

Start with agentless workspace OAuth in 2 minutes. Layer on browser extension and MCP Gateway as needed. MDM push for zero-touch rollout.

Agentless (Workspace OAuth)

Connect Google Workspace or M365 in 2 minutes. Discovers AI apps from OAuth token audit and enterprise app registrations.

Browser Extension (Chrome + Edge)

Published on Chrome Web Store and Edge Add-ons. MDM push via Chrome Enterprise policies or Intune.

MCP Gateway (21 AI Coding Tools)

Cursor, Claude Code, VS Code, Windsurf, Cline, Continue, Zed, JetBrains and 13 more. One config change per tool.
Learn more →

MDM Push (Chrome Enterprise, Intune, JAMF)

macOS .pkg with launchd, Linux .deb with systemd, Windows PowerShell installer with Task Scheduler. Zero user interaction.

Desktop Agent (Coming Soon)

Native macOS and Windows agent. Captures native app interactions that browser extensions cannot see.

get started

See what your team is actually sending

Connect a workspace read-only and you have an inventory in minutes. No agent on anyone’s machine to find out.

Book a 30-minute callRead the install guide
→ read-only OAuth, nothing installed
→ the extension and gateway are opt-in, per surface
→ every event a device can sign is signed there; the rest is recorded unsigned