---
title: AI Analysis
description: Learn how to use AI Analysis after running a report in Perspio to identify key insights, detect anomalies and generate an executive summary.
---

[Skip to content](https://help.perspio.io/help/ai-analysis#main-content)

English

Show submenu for translations

![Screenshot\_3-4-2025\_17232\_](https://help.perspio.io/hs-fs/hubfs/Screenshot_3-4-2025_17232_.jpeg?width=200&height=53&name=Screenshot_3-4-2025_17232_.jpeg)

Open main navigation

Close main navigation

- English
  
  Show submenu for translations
- [Contact us](mailto:servicedesk@perspio.io)

[Contact us](mailto:servicedesk@perspio.io)

 How can we help you?

- There are no suggestions because the search field is empty.

1. [Perspio Help Centre](https://help.perspio.io/help?hsLang=en)
2. [Reports](https://help.perspio.io/help/reports?hsLang=en)

# AI Analysis

## Learn how to use AI Analysis after running a report in Perspio to identify key insights, detect anomalies and generate an executive summary.

### Overview

**AI Analysis** in Perspio helps you explore the results of a completed report using artificial intelligence.

After running a report, the AI Analysis feature is available from the **left drawer**. It analyses the report results together with available report guidance and business context to help identify patterns, unusual activity, areas requiring further investigation, and important information that may otherwise be difficult to identify within large report datasets.

AI Analysis provides three ways to explore a report:

- **Key Insights** – identifies important patterns, key assets, and useful operational insights.
- **Anomaly Detection** – investigates unusual values, unexpected events, and potential data quality issues.
- **Executive Summary** – converts detailed report findings into a concise summary suitable for broader review or team discussions.

Completed analyses can also be revisited through **AI analysis history**, copied for use elsewhere, or downloaded as a Markdown file.

---

### Accessing AI Analysis

AI Analysis becomes available **after a report has been run**.

The feature analyses the results of the specific report run you currently have open. This means you must first generate the report results before starting an AI analysis.

 To access AI Analysis

1. Navigate to the **Reports** module.
2. Open the required report.
3. Configure any required report parameters, filters, assets, or date range.
4. Run the report.
5. Once the report results have loaded, select **Run AI analysis on the left drawer**.
6. Choose the type of analysis you want Perspio to perform.

The analysis is based on the report run that is currently available, so ensure your report filters and date range are correct before starting.

![](https://help.perspio.io/hs-fs/hubfs/image-png-Nov-03-2026-05-17-35-1624-AM.png?width=346&height=263&name=image-png-Nov-03-2026-05-17-35-1624-AM.png)![](https://help.perspio.io/hs-fs/hubfs/image-png-Nov-03-2026-05-23-18-7106-AM.png?width=221&height=300&name=image-png-Nov-03-2026-05-23-18-7106-AM.png)

 

---

### Analysis Types

Perspio provides three analysis types. Each uses the same underlying report information but approaches the data from a different perspective.

#### Key Insights

Select **Key Insights** when you want Perspio to identify the most significant information within the report.

This analysis is designed to help find the signal within a larger dataset by highlighting important patterns, relevant assets, and operational observations.

Key Insights may help identify:

- Concentrated or repeated activity.
- Assets showing significantly different behaviour from others.
- Usage or movement patterns.
- Operational trends.
- Areas that may warrant closer investigation.
- Information that is present in the report but may not be immediately obvious from reviewing individual rows.

The generated result may be presented under a heading such as **Key Findings**.

 

#### Anomaly Detection

Select **Anomaly Detection** when you want to investigate unusual or unexpected information within the report.

This analysis focuses on identifying values or events that differ from expected patterns.

It may highlight:

- Unusually high or low values.
- Unexpected events.
- Irregular activity.
- Outliers.
- Potential reporting gaps.
- Possible data quality issues.
- Behaviour that differs significantly between assets or periods.

An anomaly does not automatically indicate a fault or operational problem. It identifies something unusual enough to warrant further review.

Always compare identified anomalies with the underlying report records and relevant operational circumstances.

 

#### Executive Summary

Select **Executive Summary** when you need a concise overview of the report rather than a detailed technical analysis.

The Executive Summary converts the report findings into a shorter format that can support:

- Management reviews.
- Operational meetings.
- Customer discussions.
- Internal updates.
- Handover information.
- High-level reporting.

This option is useful when the audience needs to understand the most relevant findings without reviewing every record in the original report.

---

### How Perspio Determines the Analysis

AI Analysis uses more than the raw report rows alone.

The quality and relevance of the analysis are influenced by three main sources of context.

#### 1. The Report Run

The completed report results and available run reference determine **what is being analysed**.

This includes the dataset produced using the report configuration you selected, such as:

- Date range.
- Assets.
- Groups.
- Locations.
- Report-specific filters.
- Other parameters available for that report.

For this reason, the report should be configured correctly before starting AI Analysis.

If you run the report using an incorrect date range or asset selection, the AI will analyse those results rather than the information you originally intended to review.

 

#### 2. Report Guidance

Perspio can use guidance associated with the report to help interpret the data correctly.

Report guidance may include information such as:

- The purpose of the report.
- Field definitions.
- Metrics.
- Thresholds.
- Relevant terminology.
- Known limitations.

This context helps the AI understand what the report fields represent rather than treating the data as a generic dataset.

For example, knowing the purpose of a field or configured threshold can help the analysis distinguish between a normal value and one that may warrant investigation.

 

#### 3. Tenant Context

Business context can make the generated analysis more relevant to your organisation.

Tenant context may include:

- Business objectives.
- Industry.
- Asset types.
- Regions.
- Business terminology.
- Operational instructions.

This context helps Perspio interpret findings in a way that is more aligned with the organisation using the platform.

Two businesses may run similar reports but consider different patterns important based on how they operate.

---

### Important Considerations

AI-generated analysis should be treated as an aid to investigation and decision-making.

Always validate important findings against the original Perspio records and your organisation's operational context.

AI Analysis may identify correlations, patterns, or unusual values, but those findings do not necessarily establish causation.

Where a finding could affect safety, compliance, maintenance, customer commitments, financial decisions, or other important operations, review the supporting source data before acting on the analysis.

- [Introduction](https://help.perspio.io/help/introduction?hsLang=en#main-content)

    - [How to's](https://help.perspio.io/help/introduction?hsLang=en#how-tos)
- [Assets](https://help.perspio.io/help/assets?hsLang=en#main-content)

    - [Asset Inspector](https://help.perspio.io/help/assets?hsLang=en#asset-inspector)
    - [Utilisation Module](https://help.perspio.io/help/assets?hsLang=en#utilisation-module)
    - [Configuration](https://help.perspio.io/help/assets?hsLang=en#configuration)
    - [Events](https://help.perspio.io/help/assets?hsLang=en#events)
    - [Servicing](https://help.perspio.io/help/assets?hsLang=en#servicing)
    - [Asset Configuration](https://help.perspio.io/help/assets?hsLang=en#asset-configuration)
    - [Device Configuration](https://help.perspio.io/help/assets?hsLang=en#device-configuration)
- [Explorer](https://help.perspio.io/help/explorer?hsLang=en)
- [Dashboards](https://help.perspio.io/help/dashboards?hsLang=en#main-content)

    - [Configuration](https://help.perspio.io/help/dashboards?hsLang=en#configuration)
- [Reports](https://help.perspio.io/help/reports?hsLang=en#main-content)

    - [Available Reports](https://help.perspio.io/help/reports?hsLang=en#available-reports)
- [Groups](https://help.perspio.io/help/groups?hsLang=en#main-content)

    - [Security Groups](https://help.perspio.io/help/groups?hsLang=en#security-groups)
    - [Asset Groups](https://help.perspio.io/help/groups?hsLang=en#asset-groups)
- [Geofences](https://help.perspio.io/help/geofences?hsLang=en)
- [Workflows](https://help.perspio.io/help/workflows?hsLang=en)
- [Administration](https://help.perspio.io/help/administration?hsLang=en#main-content)

    - [Settings Module](https://help.perspio.io/help/administration?hsLang=en#settings-module)
    - [Admin Panel](https://help.perspio.io/help/administration?hsLang=en#admin-panel)
- [Devices](https://help.perspio.io/help/devices?hsLang=en#main-content)

    - [Troubleshooting](https://help.perspio.io/help/devices?hsLang=en#troubleshooting)
    - [Installation Guides](https://help.perspio.io/help/devices?hsLang=en#installation-guides)

[![](https://help.perspio.io/hs-fs/hubfs/Perspio%20Logo%20%E2%80%93%20Colour%20White%20%E2%80%93%20large%20%20RGB.png?width=786&height=177&name=Perspio%20Logo%20%E2%80%93%20Colour%20White%20%E2%80%93%20large%20%20RGB.png)](https://perspio.io)

2/50 Yeo St, Neutral Bay NSW 2089

[mailto:servicedesk@perspio.io](mailto:servicedesk@perspio.io) <https://www.facebook.com/InauroAus/> <https://www.inauro.io/>

Copyright © 2026 Inauro