← Return to All Products
HARDENED PRODUCT // PRD_6 STABLE RELEASE

LAN / Internet Chat Server

LAN / Internet Chat Server (powered by EY-Chat) is an enterprise-grade communications, real-time messaging, audio/video conferencing, and AI virtual coworker collaboration platform. Built with a Laravel 13 backend, Vue 3 client, Tailwind CSS 4, and Laravel Reverb WebSockets, the platform delivers unified communications across Web, Windows Desktop (Tauri v2), and Android Mobile (Capacitor v8) environments.

The platform supports 1:1 direct messaging with canonical ULID pairs, departmental group channels with role-based member governance, scoped message pinning ("Pin for Myself" vs. "Pin for Both"), emoji reactions, in-line quoted replies, and monotonic sequence reconciliation ensuring zero message loss and offline delivery sync. The dual calling architecture provides ultra-low latency peer-to-peer Coturn TURN/STUN WebRTC mesh audio/video calling alongside containerized Jitsi Meet multi-party video conferencing, backed by an in-browser hardware diagnostic suite (VU meters, webcam resolution check, ICE candidate gathering). Deploy Zero-Leakage AI Virtual Coworkers directly into conversation channels without sensitive data exposure, and take advantage of bilingual English/Myanmar Unicode text normalization.

Customer Flexibility: Highly customizable to fit the needs for our customers—supporting completely air-gapped private office LANs without public domains, cloud VPS deployments, custom tenant branding, and customized AI assistant tool integrations.

// HARDENED CAPABILITIES

✓ Real-Time WebSockets (Laravel Reverb & Monotonic Sync) ✓ Dual Calling: Coturn WebRTC Mesh & Jitsi Meet Conferencing ✓ Zero-Leakage AI Virtual Coworkers in Channels ✓ Cross-Platform: Web, Windows Desktop & Android Mobile ✓ Scoped Message Pinning, Reactions & In-Line Quoting ✓ Hardware Diagnostic Suite (Mic VU Meter & WebRTC Echo) ✓ Offline LAN / Air-Gapped & Cloud VPS Deployments ✓ Bilingual English & Myanmar Unicode Text Normalization ✓ Highly Customizable to Customer Needs
LAN / INTERNET CHAT SERVER
REVERB WS // AIR-GAPPED OK
Eyonics Chat Core // Secure Organization Mesh
COTURN P2P RELAY: ONLINE
#
dev-ops-cluster
WEBRTC P2P CALL IN PROGRESS:
0.0% LOSS // E2E ENCRYPTED
SC
Sarah Connor 10:41 AM DEV-LEAD

Reverb WebSocket cluster is deployed on local office LAN. Reconnecting test passed with zero packet loss.

AI
Virtual Colleague (Private vLLM) SOVEREIGN BOT
10:42 AM

Monotonic sequence synchronization verified. 48 offline messages reconciled. Private AI inference completed on local server with 0 external network calls.

👍 5 🚀 3 🔒 100% PRIVATE
10:43 AM Lead Engineer (You)

Perfect. Starting the Jitsi Meet conference room for the architecture sign-off now.

ME
Typing: Sarah Connor is typing... ENCRYPTION: AES-256-GCM ACTIVE
✓ AIR-GAPPED & CLOUD HYBRID
COTURN + JITSI MEET HIGHLY CUSTOMIZABLE
⚡

Customer Tailored: Highly Customizable Architecture

Every deployment is completely customizable to fit your exact operational requirements—from private air-gapped office LANs to cloud multi-tenant clusters with custom branding and security policies.

TAILORED DEPLOYMENT

01 / Real-Time Messaging & Sync Engine

Enterprise Messaging & Resilient State Sync.

• High-concurrency WebSockets • Zero message loss offline sync • Private LAN & WAN topologies
// CHAT ARCHITECTURE

1:1 & Group Channels

Canonical ULID-paired direct conversations and departmental group channels with role-based governance (Owner, Admin, Member), custom avatars, channel descriptions, and secure contact requests.

GOVERNANCE: RBAC ROLES ULID THREADS
// RESILIENCE

Monotonic Sequence Sync

Every message pairs an idempotent client ID with a monotonic integer sequence number. Reconnecting clients automatically reconcile missed messages, reactions, and read receipts with zero message duplication.

RECONCILIATION: SEQ-BASED POLLING FALLBACK
// MESSAGE CONTROLS

Scoped Pinning & Mentions

Choose between "Pin for Myself Only" or "Pin for Both of Us", user mentions (@username), group-wide alerts (@all), in-line quoted replies, message forwarding with original attribution, and emoji reactions.

INTERACTION: RICH MEDIA EMOJI & QUOTES
// LOCALIZATION

Unicode Text Normalization

Native Myanmar Unicode text normalization engine fixing broken vowel and consonant ordering, accompanied by bilingual English/Myanmar interface options and organizational department trees.

LANGUAGE: BILINGUAL UNICODE ACCURATE

02 / Voice, Video & Conferencing

Dual Calling Architecture: Coturn & Jitsi Meet.

• Ultra-low latency P2P mesh WebRTC • Containerized multi-party Jitsi • Floating call overlays
// DIRECT WEBRTC ENGINE COTURN MESH

Peer-to-Peer Audio & Video

High-performance WebRTC engine utilizing private STUN/TURN relays. Features SDP offer/answer negotiation, trickle ICE candidate gathering, in-call screen sharing, and a floating minimizable call overlay.

› Sub-millisecond latency on local office LANs

› Minimizable floating widget for uninterrupted chat

› High-definition desktop screen sharing

LATENCY: SUB-SECOND // E2E ENCRYPTED
// CONFERENCING ENGINE JITSI VIDE压力RIDGE

Multi-Party Video Conferencing

Integrated Jitsi Meet cluster (jitsi-web, prosody, jicofo, jvb) operating on high-bandwidth media forwarding. Supports dozens of simultaneous conference participants with prejoin bypass and secure room keys.

› Dozens of simultaneous video conference feeds

› Auto-generated secure room tokens

› Dedicated UDP port 10000 media forwarding

CAPACITY: MULTI-PARTY CONFERENCING
// HARDWARE DIAGNOSTICS ADMIN SUITE

Live Network & Hardware Testing

Built-in diagnostic console featuring microphone frequency VU meter analysis, camera resolution/framerate verification (e.g. 1080p @ 60fps), synthesized speaker chord tests, and automated ICE gathering loopback simulations.

› Web Audio API microphone volume VU meter

› End-to-end in-memory loopback call test

› Live STUN/TURN ICE candidate gathering log

DIAGNOSTICS: LIVE IN-BROWSER

03 / Enterprise AI & Topologies

Zero-Leakage AI Coworkers & Deployment Options.

• Custom virtual colleague personas • Zero external knowledge leakage • Sovereign self-hosted inference
AI COWORKERS

Virtual Colleague Personas

Provision specialized AI colleagues with custom handles, job titles, avatars, and domain instructions directly in your workspace directory.

INTEGRATION: DIRECT & GROUPS
PRIVACY BOUNDARY

Zero-Leakage Architecture

Connect to local private LLMs (Ollama, vLLM) on internal hardware or enterprise API gateways with Zero-Data-Retention guarantees.

PRIVACY: 100% AIR-GAPPED OK
CLIENT ECOSYSTEM

Desktop, Mobile & Web

Native Windows Desktop app (Tauri v2), Android Mobile client (Capacitor v8), and responsive web client all backed by unified Reverb signaling.

CLIENTS: TAURI & CAPACITOR
CUSTOM DEPLOYMENT

Private LAN or Cloud VPS

Easily deployed on a single offline office server without a public domain name, or scaled across enterprise Docker and Ubuntu bare-metal clusters.

TOPOLOGY: HIGHLY CUSTOMIZABLE

Explore Other Tooling

Other Hardened Platforms

PRD_4

Central Cyber Security Monitoring Platform

Central Cyber Security Monitoring Platform (powered by Central SIEM & its agents) is an enterprise-grade Security Operations Center (SOC) management, alert analytics, vulnerability assessment, and threat intelligence platform designed for mission-critical, multi-server infrastructure deployments. Built to overcome alert fatigue and fragmented telemetry, the platform unifies real-time host telemetry from distributed Central SIEM EDR agents (including Syscollector hardware and package tracking, continuous File Integrity Monitoring [FIM] with SHA-256 baselines, and Security Configuration Assessment [SCA] against CIS Benchmarks) with line-rate network sensors (Suricata DPI auditing 42,000+ signatures and Zeek connection state logs).At its core, a Multi-Signal Alert Correlation Engine continuously analyzes events across 8 distinct attack dimensions: agent host, source IP, Central SIEM rule ID, MITRE ATT&CK technique, authenticated user identity, malware family signature, OWASP Top 10 category, and security event type. Autonomous Level 3 (L3) Agentic AI investigation loops (AgentEngine) execute autonomous multi-turn reasoning to formulate hypotheses, gather corroborated evidence, isolate root causes, and automatically triage critical alerts (severity level ≥ 12) in under five minutes. Continuous Vulnerability Assessment synchronizes with CISA KEV and NVD 2.0 feeds, while guided CERT incident response playbooks with two-person approval governance ensure safe containment, real-time alerts to team communication channels, and boardroom-ready executive Reporting with AI certified for ISO 27001, SOC 2, and CIS compliance.Customer Flexibility: Highly customizable to fit the needs and infrastructure of our customers—from bespoke SIEM correlation rules and proprietary threat feed integration to custom-tailored CERT playbooks and localized telemetry sinks.

Deploy Product →
PRD_5

AI Integrated Penetration testing and Code Audit Platform

AI Integrated Penetration testing and Code Audit Platform (powered by Black-Orca) is an enterprise-grade web application penetration testing, vulnerability assessment, and reporting platform engineered for modern web architectures, APIs, source code repositories, and host runtimes. Built for AppSec teams, cybersecurity consultancies, and lead auditors, the platform bridges the gap between automated scanner execution, manual security engineering, AI remediation blueprints, and publication-ready client deliverables.The platform operates across 6 specialized assessment modalities: black-box web surface mapping, white-box SSH remote host runtime auditing across 11 key daemons (auditing php.ini, composer audit, world-writable directories, and daemons), OpenAPI 3.x schema fuzzing (BOLA, BFLA, Mass Assignment, and rate limiting), and multi-language SAST auditing across Git repositories and ZIP archives using SafeArchiveExtractor with Zip Slip and zip bomb defense. Black-Orca features over 50 dedicated critical CVE engines with deterministic semver verification—including zero-configuration active probes for latest critical and zero-day CVEs. A multi-provider AI remediation engine generates framework-idiomatic code fixes and verification assertions, while headless Chromium generates certified 8-section technical assessment reports and Letters of Attestation with cryptographic SHA-256 integrity checksums.Customer Flexibility: Highly customizable to fit the needs and codebase requirements of our customers—including customizable Rules of Engagement (RoE), custom CVE detection probes, framework-specific remediation blueprints, and branded executive attestation deliverables.

Deploy Product →
PRD_7

Eyonics Collaboration

Eyonics Collaboration (powered by CollabHub) is an enterprise multi-tenant collaboration platform and intelligent AI workspace. It unifies project management, cloud document storage, business communications, and autonomous AI productivity into a single coherent interface.The platform integrates real-time Kanban task management with dynamic swimlanes and subtask trees synchronized via WebSockets, an enterprise Drive document hub with native ONLYOFFICE DocumentServer 9.0 for live in-browser co-editing of Word (.docx), Excel (.xlsx), and PowerPoint (.pptx) with revision history, a multi-account IMAP/SMTP email client with vCard corporate address books, an interactive multi-view calendar with bidirectional Google, Outlook, and Zoom synchronization, real-time team chat with presence indicators, and autonomous AI Bot Agents equipped with 17 native workspace tools. Engineered with strict multi-tenant isolation (TenantScope), PostgreSQL 18 with SCRAM and verify-full TLS, TOTP 2FA authentication, and dual SuperAdmin and Tenant administration consoles.Customer Flexibility: Highly customizable to fit the needs for our customers—allowing flexible modular activation, custom tenant branding, storage quotas, custom department hierarchies, and bespoke toolchains.

Deploy Product →
High-Resolution Visual Inspection
Enlarged Visual
ESC OR CLICK OUTSIDE TO CLOSE 100% SCALE INSPECTION