Biznez Stack Documentation

Welcome to the Biznez Stack documentation. Learn how to deploy AI agents to any cloud runtime and securely connect them to 140+ MCP tool servers through our enterprise Agent Gateway.

Security First: Throughout this documentation, placeholder values (e.g., my-project, my-cluster) are used for configuration. Never paste real credentials, tokens, or secrets into documentation, tickets, screenshots, or terminals. If credentials are ever exposed, rotate them immediately.

What is Biznez Stack?

Biznez Stack is an enterprise platform for deploying and managing AI agents. It provides:

• Multi-Cloud Agent Deployment - Deploy containerized agents to GKE, Cloud Run, EKS, AKS, or local Kubernetes
• MCP Server Ecosystem - Access 140+ pre-built MCP servers (Gmail, Slack, GitHub, databases, and more)
• Agent Gateway - Secure, authenticated access to all MCP communications with OAuth 2.1 + PKCE
• Connector Management - Unified hub for LLM providers and OAuth service integrations

Architecture Overview

┌─────────────────────────────────────────────────────────────────────┐
│                        BIZNEZ STACK PLATFORM                         │
│                                                                      │
│  ┌────────────────────────────────────────────────────────────────┐ │
│  │                     AGENT DEPLOYMENT                            │ │
│  │   Your Container ──► Deploy ──► GKE / Cloud Run / EKS / AKS    │ │
│  └────────────────────────────────────────────────────────────────┘ │
│                                │                                     │
│                                ▼                                     │
│  ┌────────────────────────────────────────────────────────────────┐ │
│  │                      AGENT GATEWAY                              │ │
│  │         OAuth 2.1  │  RBAC  │  Rate Limiting  │  Audit         │ │
│  └────────────────────────────────────────────────────────────────┘ │
│                                │                                     │
│                                ▼                                     │
│  ┌────────────────────────────────────────────────────────────────┐ │
│  │                   MCP SERVER ECOSYSTEM                          │ │
│  │   Gmail │ Slack │ GitHub │ Databases │ Search │ 140+ more      │ │
│  └────────────────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────────┘

Core Concepts

Organizations & Workspaces

Biznez Stack uses a multi-tenant architecture:

• Organization - Top-level tenant boundary (your company)
• Workspace - Isolated environment within an organization (dev, staging, prod)
• Runtime - Deployment target registered to a workspace (GKE cluster, Cloud Run project)

Agent Deployment

Deploy your containerized AI agents with a unified pipeline that works across:

Agent Gateway

All MCP communications pass through the Agent Gateway for:

• Authentication - OAuth 2.1 + PKCE via your identity provider
• Authorization - RBAC at organization, workspace, and resource levels
• Rate Limiting - Protect against abuse
• Audit Logging - Complete trail of all MCP calls

MCP Server Catalog

Access 140+ pre-built MCP servers directly from Docker Hub:

Getting Started

Prerequisites

• Docker installed locally
• Access to a Kubernetes cluster or Cloud Run project
• A Biznez Stack account

Deploy Your First Agent

# 1. Login to Biznez Stack
biznez login

# 2. Create a workspace
biznez workspace create my-workspace

# 3. Register a runtime (replace with your project and cluster names)
biznez runtime register my-gke-cluster \
  --type gke \
  --project $GCP_PROJECT_ID \
  --cluster $CLUSTER_NAME

# 4. Deploy an agent (replace with your container registry)
biznez agent deploy my-agent \
  --image your-registry.example.com/my-agent:latest \
  --runtime my-gke-cluster

Note: Replace placeholder values with your actual project, cluster, and registry information.

Support

• Documentation Issues - Submit via your organization's support channel
• Community - Join our community forums
• Enterprise Support - Contact your account representative