Pergami AI is Pergamon's built-in AI support assistant. It answers product usage questions, guides you through submitting support requests, and helps you track existing tickets without leaving the application.
This article covers what Pergami AI is, what it is designed to do, its current scope and constraints, and how to get the most value from it in its Beta release.
Interface and interaction model
Pergami AI opens as a side drawer sliding in from the right side of the screen, overlaying the current page without triggering a navigation event. You remain exactly where you are in the application — no context is lost, no workflow is interrupted.
The drawer is organised across three tabs: Home, Chat, and Requests. The Home screen provides quick access to all primary actions. Chat initiates a conversation session. Requests handles structured ticket submission.
The interface is intentionally minimal and native to the Pergamon visual language — not a third-party widget overlay.
Important
Pergami AI is not a large language model (LLM) and is not designed to function as a general-purpose conversational AI. It does not support open-ended multi-turn dialogue, freeform content generation, or reasoning across arbitrary topics. It is a purpose-built support assistant with a defined retrieval and routing architecture, optimised for product usage queries, support request submission, and ticket tracking within the Pergamon ecosystem. Interactions that fall outside this defined scope will not produce meaningful results.
About the Beta release
Pergami AI is in Beta. Understanding what that means is essential to using it effectively.
Beta is not a stabilized, production-hardened release. It is a deliberate early access phase in which a working foundation is made available to real users under real conditions — specifically because lab testing cannot replicate the complexity and variability of production usage. The product is functional. It is not finished.
In practical terms, this means:
- Pergami AI will not always get it right. Responses may be incomplete, imprecise, or occasionally incorrect. This is expected behavior in a Beta release, not a defect.
- Its capabilities are intentionally constrained. Features that require greater model maturity, deeper system integration, or broader training data are scoped out of this release and will be introduced progressively.
- It does not work optimally out of the box. Pergami AI improves through usage. The more it is used and the more feedback it receives, the more accurately it can respond to the specific language, workflows, and needs of Pergamon users.
- Its behavior will change. Responses, capabilities, and interaction patterns will evolve as the model is refined. What Pergami AI can and cannot do today is not indicative of its production state.
Beta access is extended to a selected group of users precisely because their usage, edge cases, and feedback drive the development roadmap in ways that internal testing cannot. Your experience with Pergami AI in this phase is not incidental — it is part of how the product is built.
Warning
Do not use Pergami AI as the sole source of truth for critical decisions or technical troubleshooting. Validate important responses through the Pergamon Support Portal.
What Pergami AI can help with
Pergami AI is purpose-built for product support. Its current scope covers three defined areas.
Product usage questions
Pergami AI can answer questions about how to use Pergamon features — documents, assets, publications, translations, and related workflows. It draws on Pergamon's help content to surface relevant answers without requiring you to search manually.
Support request submission
Pergami AI guides you through submitting a structured support ticket. Available request categories in the Requests tab are:
- Report an Issue — something is not functioning as documented
- Suggest an Improvement — feedback on existing functionality
- Request a New Feature — capabilities you would like to see in the product
- Product Training — request training resources or a session for your team
- Translation Issue — problems with translated or localised content
Ticket status
Pergami AI can retrieve the status of a previously submitted support request. You will need your ticket number. Email verification is required to access ticket details.
Current scope and constraints
The following are deliberate constraints of the Beta release, not technical limitations to be worked around. Each represents a capability boundary that is explicitly out of scope for this phase.
| Constraint | Detail |
|---|---|
| No account data access | Pergami AI cannot read, retrieve, or reference your documents, assets, workspace files, or account-level data. |
| No product actions | Pergami AI is an assistant, not an agent. It cannot create, edit, move, publish, or delete content inside Pergamon on your behalf. |
| No multi-step troubleshooting | Guided diagnostic workflows are not available in this Beta release. For complex technical issues, submit a support request. |
| No billing or subscription support | Pricing, plan changes, invoicing, and subscription management are outside Pergami AI's scope. Contact Pergamon Support team for these enquiries. |
| No session memory | Pergami AI does not retain context between separate sessions. Each conversation begins without knowledge of previous interactions. |
| Response accuracy | Pergami AI may produce incomplete or imprecise responses, particularly for edge cases or recently updated features. Cross-reference critical answers with official documentation or your support team. |
Note
Several of these constraints are specific to the Beta phase and will be addressed in subsequent releases. The constraints table reflects the current state of the product, not its intended production scope.
How to open Pergami AI
Pergami AI is accessible from any screen inside Pergamon. Click the Help(?) icon in the top navigation bar to open the panel.
The Help icon in the top navigation bar launches the Pergami AI panel.
The panel opens with three navigation tabs at the bottom: Home, Chat, and Requests. The Home tab provides quick access to all primary actions.
The Pergami AI home screen. Start AI chat, track a request, report an issue, or book onboarding from a single view.
Using the AI chat
- Click the Help ? icon to open the Pergami AI panel.
- Click Start AI chat on the Home screen, or navigate to the Chat tab.
- Type your question in the message field and press Enter or click Send.
- Review the response. If Pergami AI cannot resolve your question, it will direct you to submit a support request.
The Chat interface. Type directly into the message field or use the quick action buttons to initiate a structured workflow.
Note
Response latency is expected behaviour. Pergami AI performs real-time retrieval and reasoning across its knowledge base before returning a response — a process that introduces a measurable processing delay. This is inherent to the architecture of retrieval-augmented AI systems and is not indicative of an error or a degraded service state. If a response does not return within a reasonable window, use the refresh icon in the panel header to reset the session.
Submitting a support request
- Click the Help ? icon to open the Pergami AI panel.
- Navigate to the Requests tab.
- Select the category that most accurately describes your request.
- Complete the guided submission flow.
The Requests screen. Select the most relevant category to route your request to the correct team.
Note
If you are unsure which category applies, select Not sure what to choose? at the bottom of the list. Pergami AI will guide you to the appropriate option through a short conversation.
Getting the best results
Pergami AI performs significantly better with well-scoped, specific inputs. Vague or open-ended questions produce lower-quality responses not because the model cannot handle complexity, but because ambiguity limits its ability to retrieve relevant context.
| Avoid | Use instead |
|---|---|
| How do I add an image? | How do I add an asset to a document in Pergamon? |
| It's not working. | I am unable to export the document |
| Tell me about translations. | How do I send a document for translation in Pergamon and track its progress? |
| What's new? | What features were added to Pergamon in the 3.9.0 release? |
Additional guidance:
- Reference the specific feature or area. Name the module or workflow you are working in. For example, "documents", "the asset manager", or "the publications workflow" to give Pergami AI the context it needs to retrieve relevant information.
- One question per message. Compound or multi-part questions reduce response precision. Submit discrete questions sequentially for better results.
- Start a new session when switching topics. Pergami AI does not carry context across sessions. Beginning a fresh conversation for an unrelated question avoids interference from prior context within the same session.
- Verify responses for critical workflows. Given the accuracy constraints of the Beta release, cross-reference Pergami AI's responses with the Pergamon Support Portal before acting on them in production-critical scenarios.
Feedback and iteration
Pergami AI's development trajectory is directly shaped by how Beta users interact with it and the feedback they submit. Incomplete responses, unexpected behavior, missing capabilities, and friction points are not just support issues, they are product signals that drive the next iteration.
There are two ways to contribute feedback:
- Inline response feedback — use the thumbs up or thumbs down on any chat response to signal response quality in real time.
- Structured feedback — go to the Requests tab and select Suggest an Improvement to submit detailed, categorized feedback to the product team.
Both channels are actively monitored. Feedback submitted during the Beta phase carries disproportionate weight in shaping the production release.
Info
For issues or enquiries outside Pergami AI's current Beta scope — including billing, subscription management, and complex technical troubleshooting contact the Pergamon support team directly through the Pergamon Support Portal.
Comments
0 comments