--- title: "Insights" url: "https://help.baramundi.cloud/en/article/insights" language: "en" last_updated: "2026-10-12T12:16:27.794Z" --- # Insights The Insights section provides you with an overview of all anomalies at the endpoints of the selected environment. The anomalies are detected and evaluated using artificial intelligence (AI). For each endpoint with anomalies, you will also receive an AI-generated analysis that provides you with troubleshooting tips. ### Overview At the top, two graphs give you a quick and easy overview of all detected anomalies: #### Distribution of anomalies detected by AI ![](https://cdn-assets.productfruits.com/dae69337-3881-425a-b803-826582591b46) Here you can see at a glance in a traffic light system (red, yellow, green) how the anomalies detected by AI are distributed across the endpoints in the selected environment. The data used for this comes from an analysis updated daily. There are four anomaly levels: - **Red:** all endpoints for which a high degree of anomaly has been detected via AI - **Yellow to orange:** all endpoints for which a moderate degree of anomaly has been detected via AI - **Green:** all endpoints with a normal anomaly level detected by AI (neither high nor moderate) - **Gray:** all endpoints with ignored status **Ignored status:** If you set the status of an abnormal endpoint to **ignored**, the anomaly level of this endpoint will no longer be taken into account or displayed in any evaluation. If the anomaly level of the endpoint changes, the endpoint automatically loses its ignored status. You also have the option of deactivating the ignored status directly on the Insights page at the endpoint. This makes the endpoint and its anomaly level visible again everywhere. #### Distribution of endpoints by anomaly level over time ![](https://cdn-assets.productfruits.com/442714b6-1f18-4da7-b651-4edd46f18674) This graph shows all endpoints with a high and moderate anomaly level over a period of two weeks. This allows you to see at a glance how the number of endpoints affected by anomalies in the selected environment has developed over the last two weeks. Improvements and deteriorations can be seen quickly and easily. The y-axis always shows you the number of affected endpoints (yellow and red together) on the respective day/date (x-axis). The tooltip shows you the total number of abnormal endpoints on that date as well as the number of endpoints with high and moderate values. ### Analysis In the analysis area, you will see a heat map on the left that shows you all abnormal endpoints, sorted by anomaly score and detection rate. On the right-hand side, you will see the details of the selected endpoint (by default, the top one in the heat map table) with the corresponding suggested solutions. #### Abnormal endpoints, sorted by anomaly score and detection rate (heat map) ![](https://cdn-assets.productfruits.com/fb4debe2-406b-451f-a72d-01f7b1c0f738) The heat map is divided into segments of equal size, which are colored according to the following anomaly levels: - Yellow: No immediate action needed - Orange: Monitoring recommended - Light red: Investigation recommended - Dark red: Immediate investigation recommended The anomaly level (colors) and the anomaly score (numbers) increase towards the top (y-axis), while the detection rate increases from 0 to 100 percent towards the right (x-axis). The heat map is interactive and shows you the selected segments in a table below ([table functions](https://help.baramundi.cloud/en/article/table-functions)). If no element is selected, you will see all affected endpoints listed in the table, sorted by default according to anomaly score (the highest at the top). As soon as you select (click on) one or more segments in the heat map, only the affected endpoints appear beneath in the table. If no endpoints are affected in the selected segments, the table is empty. Clicking on a segment again deselects it. Once you have clicked on a segment, the option to **Select all segments** appears above the heat map on the right. You can use this function to reset the selection. You also have the option of using the toggle button (switch) to **Show endpoints with ignored status**. These will then also appear in the table below. By default, ignored endpoints are not displayed in the heat map. You can change the ignored status in the details area on the right directly at the affected endpoint, also using the toggle button (switch). In the table, you will see the following columns by default: - Name of the endpoint - Username of the endpoint user - Anomaly level (high = red, moderate = yellow, normal = green (the green status does not appear in the heat map and its table)) - Anomaly score: the anomaly score is used to classify the endpoint according to its anomaly level (0 - 0.89 normal, 0.90 - 1.00 moderate, >1 high) - Trend anomaly score (number of increases or decreases compared to the previous detection) - Detection rate (percentage indicating how often a noticeable anomaly score has been recorded for this endpoint in the last seven days) - Anomaly category: the area from which the anomaly originates; If there are several, this is indicated by a small number and a plus sign. Available categories are: Stability, Hard drive, CPU, RAM, System, Battery, Network. - Last detection (time when this endpoint was last detected; the last contact with the endpoint) - Status (abnormal, ignored) (Note: The ignored status is only visible in the table if you have previously activated the **Show endpoints with ignored status** toggle button (switch) above the heat map.) ### Analysis Details #### Insights: AI-generated analysis In the detailed AI-generated analysis area, you will always see the endpoint selected on the left with its anomalies in detail. By default, the first one from the heat map table is displayed. ![](https://cdn-assets.productfruits.com/1f458410-1d5c-479e-a5a6-0e7bb910f9a1) **Ignored status:** Use this toggle button (switch) to set the status of an endpoint to ignored or back to the default non-ignored status. If you set the status of an abnormal endpoint to **ignored**, the anomaly level of this endpoint will no longer be taken into account or displayed in any evaluation. If the anomaly level of the endpoint changes, the endpoint automatically loses its ignored status. You also have the option of deactivating the ignored status directly on the Insights page at the endpoint. This makes the endpoint and its anomaly level visible again everywhere. **Actions** button: This button gives you the following options: - **Open** the endpoint (in the same window) - **Open** the endpoint **in a new tab** - Open the endpoint with **Open in baramundi Management Center** (only for endpoints that have a baramundi Management Agent installed or are managed via the baramundi Management Suite), see also [Endpoints](https://docs-staging.baramundi.com/helpsetid=m_t_usability-and-ui&externalid=proactive-hub_endpoints&Language=en-us) - **Connect via baramundi Remote Desk** to the endpoint (only for endpoints that have a baramundi Management Agent installed or are managed via the baramundi Management Suite), see also [Endpoints](https://docs-staging.baramundi.com/helpsetid=m_t_usability-and-ui&externalid=proactive-hub_endpoints&Language=en-us) - Forward insights: When you click **Forward insights**, a dialog box opens that allows you to edit the **Subject** and enter **Additional information** at the bottom. The default subject line is "Insights: AI detected analysis". The information that will be sent is also listed: Finally, click **Send** if you want to forward the AI analysis of the selected endpoint to the specified destination address. - - Who is sending this email: The sender's email address is entered here. - From which category of baramundi Proactive Hub does the data originate? - What is the name of the endpoint from which the data originates? - Overview of the most important data relating to the anomaly: Anomaly level, anomaly score, trend anomaly score, detection rate, status. - Preview of the AI-generated analysis **Anomaly level:** The anomaly level is displayed graphically for the selected endpoint. The classification is shown by the gray marker with the exclamation mark (normal, moderate, high). **AI-generated analysis:** In the AI-generated analysis section, you will see the top three anomalies from the most notable anomaly measurement time point in the last seven days. If there are fewer anomalies, you will see fewer than three results. The AI-calculated results are divided into: - A meaningful **title** - The associated **metric** (technical performance indicator) - The associated **anomaly category** - **Investigate button**: If available, use it to navigate directly to the location associated with the anomaly. - **Impact on the overall score** (based on the anomaly score in percent); Shows, as a percentage, the extent to which the respective abnormality is factored into the assessment. Higher values indicate a greater deviation from the normal state. Since multiple abnormalities can contribute to the result, the abnormality with the highest influence value is not necessarily the primary cause. - **Troubleshooting checklist:** Here you will find a practical checklist that points out the most important troubleshooting points. - **Explanations:** - - What it measures: The AI explains what was measured in this anomaly and what the measurement means. - How it works: Here you will find background information on how the system works. - Example: A simple example is used to illustrate the effect of the anomaly. - What it can lead to: AI is used here to list a selection of possible consequences and effects of the anomalies on the endpoint.