Overview
Holaboss is the infrastructure an AI-native company runs on — AI employees, the channels they work in, and the portal your customers sign in to.
Getting started
Holaboss is the infrastructure an AI-native company runs on. In one place you hire AI employees, give them tools, skills, and channels to work in, connect the data they need, and serve your customers through a branded portal — with billing, memory, and observability built in.
The pieces
- Employees — the agents that do the work. Each has a mandate (its standing instructions), a model, a set of abilities, and a memory that carries across runs.
- Channels — where an employee works: Slack, WeChat / WeCom, Discord, or an embeddable web widget. You connect a channel once and the employee shows up there.
- Skills & tools — capabilities you turn on per employee — web search, code execution, MCP integrations, and reusable skill packs.
- Your App (the portal) — the customer-facing site your clients sign in to: their orders, files, deliverables, and a thread with your team. Built from a template, or your own app.
- Billing & credits — usage metering and prepaid credits are built in; every model call is attributed and debited automatically.
- Observability — every agent run is traced, so you can see what a model was asked, what it returned, and what it cost.
Where to go next
- Bring your own portal — build the customer-facing portal on
your own stack with the publishable key and the
@holaboss/clientSDK.
This documentation is being rebuilt around Holaboss as it is today. Sections will fill in over the coming updates.