The 5 Classes of Agentic AI Solutions
Not all enterprise agentic AI is the same. The right class of solution depends on scope, governance needs, and how much judgment the AI needs to apply.
Choosing the Right Solution Class
Not all enterprise agentic AI is the same. The right class of solution depends on scope, governance needs, and how much judgment the AI needs to apply. These five classes form a spectrum from narrowly scoped personal tools to always-on enterprise agents.
Personal Productivity
Automates repetitive individual tasks within a single user's context and permissions.
Conversational Business Process
Guides employees through structured multi-step processes using natural language.
Interactive MCP App
Exposes enterprise MCP tools through deterministic forms and dashboards.
Autonomous Agent
Receives a high-level goal and executes across multiple enterprise systems.
Digital Worker
Always-on agent with persistent memory that adapts to individual user patterns.
How to Choose
| Solution Class | Scope | User Interaction | Governance Complexity |
|---|---|---|---|
| Personal Productivity | Single user | On-demand | Low |
| Conversational Business Process | Department | Guided conversation | Medium |
| Interactive MCP App | Department | Form/dashboard UI | Medium |
| Autonomous Agent | Multi-system | Goal-activated | High |
| Digital Worker | Multi-context | Always-on | Very High |