Your new AI workplace: 10 helpful functions
Welcome to the future of visual collaboration.
You already know DEON as a powerful platform for your 'heavy load' of data and visual collaboration. With the new AI integration (DHI) and the DAIA assistant (DEON AI Assistance), your workspace transforms from a passive work surface into an intelligent partner.
This overview presents you with 10 helpful functions for your daily work.
By the way: You can find an overview of the DAIA control elements here:
1. Semantic Project Query (Knowledge Retrieval)
What is it?
You can ask the DAIA assistant questions about the content of your entire DEON project. The AI not only analyzes text but also understands the context of unstructured data collections that you have stored on the board.
What is it good for?
It saves massive research time. Instead of rummaging through hundreds of documents, email screenshots or notes, simply ask the AI for the information you are looking for.
Practical example (HR):
An employee changes departments and has unloaded their knowledge (handover protocols, email exchanges, process diagrams) unstructured onto a DEON board. A new Key Account Manager simply asks: "What are my tasks and responsibilities as a Key Account?" and receives a precise summary from the scattered information.
Operation:
- Open the DAIA Panel (on the right in the menu).
- Make sure that the project has been analyzed (click on the "brain" symbol if it is red or if DAIA prompts you to do so).
- Type your question in the chat field and send it.
2. Visual Grounding (Source Referencing)
What is it?
The AI not only provides an answer but also delivers interactive reference buttons (links) to the exact locations within the project.
What is it good for?
It prevents 'hallucinations' and builds trust. You don't have to rely on the black box; instead, you can verify every piece of information immediately in its original context.
Practical example (Technical Development):
You ask about the price of a specific component (e.g., 'What is the cost of the AS5600?'). The AI provides the price. By clicking on the reference, DEON immediately zooms to the location on the board - for example, in an embedded Excel spreadsheet or a PDF specification - where this price is listed.
Operation:
- Ask a question in the DAIA Panel.
- Click on the reference buttons in the AI's response below.
- The Workspace automatically zooms to the corresponding location.
3. Deep Web Content Analysis
What is it?
DEON's AI kernel can read, understand, and incorporate content from embedded web browsers (e.g., MS Web Apps, Google Sheets, Jira, intranet pages) into its responses.
What is it good for?
You can ask questions about data that is not physically stored in the DEON project but is linked via a browser. This connects silos.
Practical example (product development):
You have embedded a web table with bill of materials as a web browser in DEON. You ask the AI: 'Which component is the most expensive on the list?'. The AI extracts the data live from the browser view and answers the question without you having to import the file.
Operation:
- Drag and drop a URL (e.g., Jira board or MS Web App) into the workspace or paste it from the clipboard (Ctrl+V).
- Ask a question in the DAIA panel that relates to the content of this webpage.
4. Visual Viewport Analysis ("The Eye")
What is it?
It allows the AI to "see" exactly what you have on your screen (viewport). The AI analyzes the visual image, including diagrams, sketches, or technical drawings.
What is it good for?
Perfect for understanding complex technical representations, diagrams, or unlabelled graphics.
Practical example (engineering):
You zoom in on a photo of a complex test setup (e.g., a test rig for actuators). You activate the "Eye" and ask: "Explain this setup." The AI describes the components (Load Cell, Actuator, etc.) and their function based on the image and considering the further knowledge within the project.
Operation:
- Zoom in on the area you want to analyze in the Workspace.
- Click on the eye icon in the DAIA panel.
- Enter your prompt (e.g., 'Explain that').
Collaborative AI Discussion
5. Graph Chat (Discussion in Workspace)
What is it?
The chat history with the AI is freed from the narrow sidebar window and mapped as a visual network of text nodes and connection lines directly on the infinite, collaborative DEON workspace.
The key benefits:
- Overview instead of 'scrolling wasteland':
While conventional chat tools lose track of long discussions, you use the zoomable DEON workspace here. You can zoom out to see the entire argument chain or zoom in to read details. - From chat to content (seamless further processing):
The chat bubbles are not static text, but fully editable, native DEON elements. You can directly format, shorten, combine, or group the AI's responses with other content. - Collaboration:
The conversation doesn't take place in a private window, but on the board. If you share the board with others, the entire team can follow the discussion live. - Present results directly:
Since the chat nodes are DEON objects, you can easily mark the result of a research and export it immediately as an editable PowerPoint slide or PDF. - Persistence:
Your graph chats remain in the project. You can continue the conversation at any point in the tree structure.
Practical example (Market research & Reporting):
A team is working on market trends for 2025. Instead of laboriously copying the AI responses from a chat window into a presentation, they let the discussion emerge as a graph on the board. The team deletes irrelevant nodes, highlights the most important findings in color, adds images, and exports the final strand directly as a finished slide.
- Enable Graph Chat: Click at the top of the DAIA panel on the Graph-Chat icon. Your ongoing or new conversation now appears automatically as a chain of text boxes directly on the board.
- Process results further: Since the chat bubbles are native DEON elements, you can freely move them via Drag & Drop, reformat them or add further content.
- Export: Select the desired chat nodes on the board and choose in the context menu (right-click) “Share/Export“, to convert the discussion thread directly into an editable PowerPoint file or a PDF.
6. Conversation Branching
What is it?
You can branch off at any point of an existing graph chat “ (create a branch). Simply click on an older answer in the graph and ask a new question – a new branch visually grows from the discussion tree.
The key benefits:
- Perfect 'Prompt Hygiene' (Context Purity):
When you change the topic in a linear chat, you "dilute" the AI's context (the Context Window). With branching, the main thread remains clean and focused, while you clarify side issues in isolated branches. - Knowledge trees instead of linear threads:
You can systematically untangle complex problems. One branch delves into technical details, another clarifies legal questions, a third addresses costs – all starting from the same point. - Quick Context Jumps:
By simply clicking on a node in the graph, the chat panel immediately jumps back into the context of this specific conversation branch.
Practical Example (Technical Development):
An engineer discusses the functionality of a 'Load Cell' in the main thread. The detailed question 'What material is the strain gauge made of?' is clarified in a side branch. After this is done, he clicks back to the main path and continues to discuss the signal electronics undisturbed.
- Select starting point: Click on any node (a previous question or answer) within your existing graph chat directly on the board. The DAIA panel now jumps exactly to that point in the conversation flow.
- Create branch: Type your new detailed question in the DAIA panel and submit it. From the selected node, a new branch now visually grows for this conversation thread. Alternatively, you can also select a previous question and modify it to create a new branch.
- Jump between strands: To return to your original main strand or another subtopic, simply click the corresponding node in the respective branch on the board. The chat panel instantly switches context and you can continue discussing there undisturbed
AI-supported Productivity
7. Integration of agent systems
DEON acts as a bidirectional visual interface for external agent systems and automation platforms such as n8n, Zapier, Make or autonomous AIs. Through an optimized REST API as well as the token‑saving DEON CLI tool combined with text‑based skills (SkillsMD) these systems can be seamlessly integrated into DEON. External agents can thus query unstructured knowledge from DEON projects (via Retrieval API for AI grounding) and autonomously place and modify results in the form of texts, images or color markings directly in the workspace.
It bridges the gap between automated AI background processes and human decision-making. Instead of painstakingly aggregating data manually from databases or logs, DEON transforms through the connected agents into a self-updating situational view (“Self-Building Dashboard“). The autonomous systems do the preparatory work and curate complex data visually, while you as a user retain full overview and decision control (“Human in the Driver Seat“).
An external monitoring system reports the failure of a server component. A connected n8n agent workflow is triggered and uses the DEON API to precisely mark the affected server symbol in the architecture diagram on your DEON board in red. At the same time, the agent places the relevant error logs as text cards right next to it. When your team enters the virtual “War Room“ (the DEON board), the error diagnosis is already visualized and everyone can immediately collaborate on the solution.
- Authentication: Generate in your DEON account management a permanent API token and store it securely as an environment variable (Bearer token) in your agent's system.
- Define target elements: Via the context menu of an object in the DEON workspace you can copy the exact project ID and element ID under “API URLs“ with a single click.
- Execute actions: Pass these IDs to the HTTP nodes of your agent (e.g., in n8n) or use the local DEON CLI tool so that the agent can query, overwrite, or create new elements in your project precisely
8. Generative Content Creation (Image & Text)
What is it?
Creation of new content (images or text) directly on the board, based on prompts or existing elements.
What is it good for?
Rapid prototyping, visualization of ideas, or creation of variations.
Practical example (design/marketing):
You have an image of a robot on the board. You select the image and tell the AI: “Create a similar image, but with more blue accents”. The AI generates the new image and places it right next to the original.
Operation:
- Select an element (image or text) in the workspace.
- Enter your request in the DAIA panel (e.g., “Create a variation of this” or “Translate this”).
- The result is placed in the workspace.
9. Semantic Context Recognition
What is it?
The DEON AI kernel understands that things that are visually close together belong together, even if they are not formally grouped. It reads the “layout” like a human.
What is it good for?
You don't have to structure your boards perfectly for the AI to understand them. Free working is supported.
Practical example (Management):
You have placed a heading “Goals until 2030” loosely over a textbox. The AI automatically understands that the content of the textbox are the goals until 2030, even if the word “Goals“ does not appear in the separate textbox below.
Operation:
This works automatically in the background. You just need to arrange your content logically on the board, as you would for a human viewer.
10. Export of AI results
What is this?
Since all AI interactions (Graph Chat, generated texts) are native DEON elements, they can be formatted and exported directly into editable formats (e.g. PowerPoint, PDF or ODP).
What is it good for?
Seamless transition from analysis to presentation. No more copy-paste chaos.
Practical example (reporting):
You have developed a complex analysis with the AI, which is displayed as a graph chat on the board. You draw a frame around it, quickly reformat the boxes, and export the selection as PowerPoint. You receive an editable slide with the results for the management meeting.
Operation:
- Select the AI results (e.g., the chat tree) on the board.
- Click on Share/Export in the context menu.
- Select the desired format. The elements remain editable in PowerPoint (not a screenshot!).
DEON Holistic Intelligence: Your workplace that thinks along.
Learn more: