Home Products About Contact
Log In / Sign In
Products

Three products, one AI operating loop

Launcher is the desktop AI entrance for staff. Central is the marketplace and operations portal. Xbase is the model gateway and compute control layer.

Launcher enterprise AI work scenarios
Launcher
Launcher
Entrance · AI task control centre
Desktop AI Entrance

Launcher — Multi-Agent workbench for enterprise staff

Launcher gives business users one practical place to chat, create, analyse, and execute tasks with approved Agents and Skills. It is designed for real office workflows, not demos.

  • ✓ Multi-agent switching for different work scenarios
  • ✓ Direct Skills calls from Central's approved catalogue
  • ✓ Cron-based scheduled task automation
  • ✓ Long-term memory across sessions
  • ✓ Local-first handling for sensitive content
  • ✓ Policy routing across local, dedicated, and shared compute
  • ✓ Multi-channel: WhatsApp, Feishu, WeCom, WeChat, 5G Messaging
  • ✓ Usage tracking, appearance themes, and auto-update
Central vertical AI applications marketplace
Central
Central
Marketplace · Vertical applications
Marketplace & Operations Portal

Central — Managed catalogue of Agents, Skills, and tenants

Central organises approved Agents and Skills into a managed catalogue. Enterprises can publish, assign, monitor, bill, and improve AI capabilities across departments and tenants. It includes a web portal for resource operations and a dedicated Console for enterprise decision-makers.

  • ✓ Agent Store and Skills Library with managed approval
  • ✓ Tenant isolation and quota control
  • ✓ Usage statistics, audit trails, and billing
  • ✓ Web portal for resource operations and admin
  • ✓ Dedicated Console for enterprise bosses and member management
  • ✓ Partner, reseller, and solution-provider distribution
Xbase AI infrastructure layer
Xbase
Xbase
Infrastructure · AI base and compute control
AI Infrastructure

Xbase — Unified MaaS gateway and compute control

Xbase provides the model service gateway and compute control layer that supports both Launcher and Central. It manages providers, tokens, quotas, routing, and audit logging with a unified API across GPU, CPU, and domestic chip resources.

  • ✓ Unified MaaS API with provider and model registry
  • ✓ Token-based access control with quotas and audit
  • ✓ GPU, CPU, and domestic chip resource management
  • ✓ Intelligent inference scheduling and load balancing
  • ✓ Shared, dedicated, private cloud, and on-prem deployment
  • ✓ Training, fine-tuning, and inference lifecycle support
How It Works

Demand, capability, and compute reinforce each other

The loop starts with staff using Launcher's multi-agent chat. Usage data drives better Skills, stronger Agents, tenant policies, and smarter compute planning in Central and Xbase.

Step 1

Open Launcher Entrance

Users start AI tasks from desktop or mobile.

Step 2

Policy Routes Work

Requests go to local, dedicated, or shared execution.

Step 3

Central Assigns Capability

Approved Agents and Skills are discovered and invoked.

Step 4

Xbase Executes

Model services and compute resources complete the task.

Step 5

Operations Improve

Usage data informs Skills, Agents, cost, and capacity planning.

Want to see the product loop in action?

We can walk through the AI Entrance, vertical application marketplace, and infrastructure setup based on your sector and deployment constraints.

Request a Demo About LCX-Asia