Limitations
The Open iT Generative AI assistant provides natural language support for interacting with Open iT tools and services. However, it operates within defined boundaries. The following sections present AI assistant's known limitations, grouped by relevant areas, to guide you in understanding where it may require additional input, external tools, or expert involvement.
Knowledge and Scope Limitations
These limitations define the AI assistant’s scope of expertise, which is restricted to Open iT tools and documentation.
| Usage Context | Limitation |
|---|---|
| Open iT-specific knowledge | Cannot assist with unrelated third-party software, general IT topics, or non-Open iT technologies. |
| Limited awareness of future features | Cannot provide information on unreleased features, roadmap items, or undocumented updates. |
Functional and Technical Boundaries
These limitations describe what the AI assistant can and cannot do technically, including system integration and tool dependencies.
| Usage Context | Limitation |
|---|---|
| No integration with external systems | Cannot execute commands or perform actions on non-Open iT platforms such as external SQL tools or Power BI. |
| Input dependency for tool-driven features | Requires structured input like reference IDs, field names, or filters to complete queries or visualizations. |
| Data retrieval limit | Can only retrieve up to 100 rows per query; larger data sets require multiple queries. |
Interpretation and Decision Support
This section highlights the importance of clear input and the AI assistant’s role as an advisory, not decision-making tool.
| Usage Context | Limitation |
|---|---|
| Requirement for clear and specific prompts | May return incomplete or irrelevant responses if queries are vague or unclear. |
| Advisory function only | Does not take action or make decisions; final decisions must be made by administrators. |
Visualization and Reporting Limits
These limitations focus on how the AI assistant handles data visualization and report generation using Open iT templates.
| Usage Context | Limitation |
|---|---|
| Structured data requirement for visualizations | Cannot create graphs or charts without well-formatted and complete data input. |
| Use of predefined report templates | Cannot create new templates or significantly modify existing SSRS RDL templates. |
Analytical Constraints
The AI assistant helps analyze Open iT data but does not perform advanced or predictive analytics.
| Usage Context | Limitation |
|---|---|
| No predictive analytics capabilities | Cannot forecast trends or use machine learning techniques for future prediction. |
| Analysis based solely on Open iT data | Does not incorporate external business context, benchmarks, or third-party datasets. |
Security and Privacy Considerations
This section addresses the AI assistant’s access restrictions and adherence to Open iT’s permission framework.
| Usage Context | Limitation |
|---|---|
| No access to confidential customer data | Cannot access sensitive or proprietary data unless explicitly provided via secure and permitted channels. |
| Adherence to user permissions | Respects Open iT role-based access controls; cannot bypass user restrictions. |
General Generative AI Behavior
These are standard limitations inherent to generative AI technologies that may impact output quality or timeliness.
| Usage Context | Limitation |
|---|---|
| Potential delay in reflecting new features | May not immediately reflect the latest Open iT updates until documentation is integrated. |
| Potential for inaccurate or incomplete outputs | Responses may require validation, particularly for compliance, financial, or technical use. |