Differences, scenarios, and decision-making aids for selecting the right agent type
The choice between a standard AI agent and a specialised AI agent depends on the use case. The most important differences, possible applications and decision-making criteria are described below.
- Overview of agent types
- Comparison of features
- Use cases for standard AI agents
- Use cases for specialised AI agents
1. Overview of agent types
Two agent types are available:
- Standard AI agents use a general knowledge base.
- Specialised AI agents offer additional control options through instructions, resources and AI actions.
2. Comparison of features
|
Feature |
Standard AI agent |
Specialised AI agent |
|
Answers |
Pure instant response |
Instant response & follow-ups (e.g. forms) |
|
Database |
Central, general knowledge base |
Customised instructions & additional sources |
|
Actions |
General backend calls such as webhooks |
AI actions (e.g. product search, retrieve tariffs) |
|
Follow-Up capability |
❌ Not recommended |
✅ Suitable, e.g. cancellation or application forms |
3. Use cases for standard agents
Standard AI agents are suitable for:
- General FAQs: ‘How much is the basic fee?’
- Static information: Company address, opening hours, guidelines.
- Simple pricing information: Reading tables without additional processes.
4. Use cases for specialised AI agents
Specialised AI agents are recommended for:
- Process forms: Cancellation requests, damage reports, meter readings
- Real-time data queries: Product search with price and stock information
- Advanced instructions: Subject-specific rules, e.g. for financial advice, exclusion of sensitive topics
- Supplementary resources: Product catalogues, contact lists, tariff overviews
Decision aid
The choice of agent type depends on the use case. Standard AI agents are suitable for simple information. Specialised AI agents are recommended if additional instructions, process steps or AI actions are required.
The creation of AI agents is described in this article.