The Qlik Cloud Discovery Agent – Part 3: Get the most out of your Insights feed

In part 1 of this series, we set up the Discovery Agent from the administrator's perspective. In part 2, we looked at how developers create insight triggers that deliver meaningful results. In this third and final part, we shift our focus to what matters most day to day: you, the end user checking the feed.

The feed is where the Discovery Agent proves its value. Everything configured in parts 1 and 2 comes together here in a scrollable list of insight cards that tell you what has changed in your data, where it happened and why it matters. This article covers how to find the feed, how to read it, how to filter it down to what matters to you and how to dig deeper when an insight calls for it.

Where do you find your feed?

The Insights feed is accessible from the Insights or Analytics activity center in Qlik Cloud. In the menu on the left, click Feed. That is all the navigation you need.

Your feed is personal. You only see insights from applications you have access to, and only for the measures that have insight triggers configured. If the feed looks empty, it is either because no triggers have been created yet in the applications you use, or because no reload has taken place since the triggers were created. Both are worth checking with your application developer if you expected to see content.

Reading an insight card

Each insight appears as a card in the feed. Every card is designed to answer three questions at a glance: what changed, where it happened and why it matters. The first two are facts taken straight from the data. The third is context generated by the Discovery Agent to help you understand the significance of what you are looking at.

Cards show the name of the measure, the time period in which the change occurred and the nature of the detected change: a spike, a new record, a trend reversal and so on. You can expand a card for more detail, including a small chart showing the historical pattern and the specific data points that triggered the insight. Each card is also linked to the application it came from, so you can navigate straight from the feed to the full dashboard if you want to investigate further.

Please note: insights in the feed expire after 90 days. If an insight catches your attention but you cannot act on it right away, make a note of it or open the underlying application, as you cannot save it for later.

Filtering your feed

In a well-configured environment with multiple applications and many active insight triggers, the feed can quickly get busy. The filter options at the top of the feed are your main tool for keeping it manageable.

You can filter by:

• Measures: show only insights for a specific metric, such as revenue or order count
• Time dimensions: focus on insights related to a specific date field
• Time period: show only insights for a particular aggregation period, such as daily, weekly or monthly
• Field values: narrow down to a specific value within a breakdown dimension, such as a particular region or product category
• Insight type: filter by the kind of change detected, such as spikes, records or trend changes

These filters can be combined. If you are responsible for the European sales region and only want to see weekly revenue insights, set those two filters and the feed becomes much easier to follow.

Digging deeper with Qlik Answers

When you want to understand an insight card better, you can open it directly in Qlik Answers without leaving the feed. The context of the insight card, including the application, the measure and the relevant time period, is automatically passed on to Qlik Answers so you can start asking follow-up questions right away.

For example, if you see a spike in customer returns last week and want to understand whether it is concentrated in a particular product category or region, simply ask that question in plain language. Qlik Answers uses the underlying data of the linked application to respond, with references so you can trace the answer back to its source.

It is worth understanding exactly what Qlik Answers can access here. When you open Qlik Answers from an insight card, it works with the structured data of the linked Qlik application. It does not automatically include any separately configured knowledge bases with unstructured content. If your organization uses Qlik Answers assistants with document-based knowledge for broader questions, those are available through the Qlik Answers interface, not from the feed.

Above model / Below model

These two types are the most advanced. Rather than comparing values to a simple historical average, the Discovery Agent builds a predictive model based on the historical data and evaluates whether the actual values turn out significantly higher or lower than that model's prediction.

The model takes patterns in the data into account when generating predictions, including underlying trends and recurring cycles. This makes above model and below model particularly valuable for measures with seasonal behavior. A spike detection algorithm might flag a peak in retail sales in December as unusual without realizing that December is always high. A model-based approach builds that seasonal pattern into the prediction and only triggers an insight when December is exceptionally high, even by December standards.

The minimum data requirement is 4 data points for non-daily aggregations and 7 for daily ones, the same as for spike detection. The model can be built on a relatively small amount of data, but more history naturally produces better predictions. If your data has a strong seasonal component and you notice that spike detection produces too much noise, switching to above/below model is the logical next step.

Data volume requirements at a glance

Before you create your first trigger, it is worth understanding the minimum data requirements. Creating a trigger on a measure with insufficient data does not produce an error, but it will never generate any insights either. That can be hard to diagnose if you do not know what to look for.

When are insights generated?

The Discovery Agent evaluates insight triggers at most once a day, whenever the application data changes as a result of a reload. If an application is reloaded several times a day, only the first reload of that day triggers the insight evaluation.

There is an important distinction between the first evaluation and all subsequent ones:

• First evaluation: when you create a trigger and the data is updated for the first time, insights are calculated over the last seven data points. For a daily trigger, that means seven days; for a monthly trigger, seven months. Historical data beyond those seven points can be used as analytical context, but insights are only generated for the most recent seven.
• Subsequent evaluations: after that first update, the Discovery Agent only looks at data points that are new since the last evaluation. If you reload daily, each daily reload only generates insights for that day's new data.

This has a practical implication: if you create a trigger today on an application that has been running for years, your users will not immediately see a backlog of historical insights. They will see insights for the last seven data points, after which new insights appear as fresh data comes in.

One more timing detail worth knowing: current and future time periods are always excluded from insight calculations. A monthly trigger does not generate insights for the current month until that month is over. This makes sense, but it is easy to forget when you are testing a newly created trigger and wondering why nothing shows up.

Tips and tricks

Start with spike detection

If you are setting up triggers for the first time, start with Spikes up and Spikes down. They have the lowest data volume requirements and start generating insights quickly. Once you have confirmed that the pipeline works end to end, you can gradually add more insight types.

Check your date fields before you start

Before creating triggers, check that the date fields you want to use as a time dimension are actually loaded as date values in the data model. Open the data model viewer, look up the field and check that its type is shown as date or timestamp rather than as a text string. If it is a text string, update your load script using the date() or timestamp() functions. This is one of the most common reasons why a newly created trigger produces no insights, and it can be hard to diagnose if you do not know what to look for.

Be deliberate in your choice of insight types

Enabling all eight insight types for every trigger feels thorough, but it results in a busy feed. Think about what each measure actually means. For a stable operational measure such as average order value, spike detection and above/below model make sense. For a strategic KPI such as net promoter score, change in baseline and change in trend are probably what you are looking for. Choose the types that match the question you are actually trying to answer.

Use breakdown dimensions strategically

Breakdown dimensions multiply the number of calculations and the number of insights in the feed. Adding a breakdown dimension with 50 product values means every evaluation now performs 51 calculations: 50 individual values plus the overall aggregate. That is powerful, but only if the product-level breakdown is genuinely useful to your users. Start without breakdown dimensions, let users work with the aggregated insights first and add breakdowns once you know which dimensions they actually want to filter on.

What publishing and duplicating do to your triggers

Insight triggers live in an application, but they are not carried over in every scenario. When you publish an application to a managed space, the triggers have to be recreated in the published version. The same applies when you duplicate an application, or export and re-import it: triggers do not survive duplication.

Moving an application between spaces is the only scenario in which triggers are preserved. Keep this in mind when planning your development and deployment workflow, especially if you maintain a development version of an app that you periodically publish to production.

Section access is a hard blocker

If your application uses section access for row-level security, insight triggers are not available for that application at all. This is a fundamental limitation of the current version of the Discovery Agent, not a configuration issue you can work around. Take this into account if section access is a requirement in your environment.

In the next part

Now that your insight triggers are configured and your reload schedule is in place, the Discovery Agent will start populating your users' feeds after the next data update. In part 3 of this series, we switch perspective to the end user and look at how to read and navigate the feed effectively, how to use filters to cut through the noise and how to use Qlik Answers to dig deeper into an insight that catches your attention.

To use Qlik Answers from the feed, you need two separate permissions set up by your administrator. The Insights feed permission (under Features and actions > Agentic AI) gives you access to the feed itself. The Data analysis permission (also under Features and actions > Agentic AI) is additionally required to use Qlik Answers. If the option to explore further in Qlik Answers from an insight card is not available, ask your administrator whether both permissions are enabled for your account.

This combination of proactive detection and conversational exploration is at the heart of Qlik’s agentic analytics experience. The Discovery Agent tells you what is worth looking at; Qlik Answers helps you figure out what to do about it next.

Tips and tricks

A quiet feed is not a broken feed

The Discovery Agent does not show every change in your data. It evaluates statistical significance and only surfaces insights for changes that genuinely deviate from the normal pattern. If the feed is quiet on a given day, the system is working exactly as intended: it means your metrics are moving within expected limits and nothing currently requires your attention.

If the feed stays empty for a longer period in an application you know has active data, that is the time to check with your developer whether the triggers are configured correctly and whether the data has enough history for the insight types being used.

Manage a busy feed with filters

When the feed is first activated for an application, or after a new trigger has been created, you may see more insight cards than you would normally expect. The first reload after a trigger is created performs a one-off historical scan over the most recent data points, which is why that first run usually produces more cards than usual. After that first update, each reload only surfaces insights based on newly arrived data.

Rather than trying to read every card, use the filters to focus on the measures and dimensions you actively manage. A filtered feed that is relevant to your role is far more actionable than an unfiltered stream of everything happening across all applications.

Reload frequency sets the pace of your feed

Insights only appear after a data reload. If an application is reloaded once a day, your feed for that application is updated once a day. If it is reloaded weekly, you will get weekly insights at most.

This is useful to know when setting expectations with your team. If you want to monitor a measure daily, the underlying application also needs to be reloaded daily. If your current reload schedule does not match how often you want to receive insights, discuss it with your application developer or data team.

Use Qlik Answers for the follow-up question

An insight card tells you that something has changed. It does not automatically tell you why it happened or what you should do about it. That is where Qlik Answers comes in. Do not just use it to confirm the numbers, but to ask the follow-up questions: which customer segment caused the spike? Which product made the margin drop? In which region is the trend change occurring?

The more specific your question, the more useful the answer. The context passed on automatically from the insight card gives Qlik Answers a head start, but asking precise, targeted follow-up questions leads to a usable conclusion faster than broad, open-ended ones.

Do not use the feed as an archive

Insights expire after 90 days. If an insight is important enough to record, export the relevant data from the application or document the finding in your regular reporting. The feed is intended for timely awareness and action, not for long-term archiving.

What the future holds

The Discovery Agent is still a relatively new capability within Qlik Cloud, and there are a few current limitations worth knowing about. The most significant for many organizations is section access: applications that use section access for row-level security currently cannot generate insights in the feed. Support for section access is on the roadmap for a future release, so if this applies to your environment, it is wise to keep an eye on the Qlik release notes.

Several additional capabilities are also planned. A Following tab that lets you follow specific applications or insight categories is on its way, as are insight delivery by email and integration with external tools. If the current feed-based experience is already adding value for your team, the roadmap suggests it will become even easier to use in the future.

Wrapping up the series

This third article concludes our series on the Qlik Cloud Discovery Agent. Across three parts, we have covered the complete picture: how an administrator enables the capability, how a developer builds effective insight triggers and how an end user gets the most out of the resulting feed.

The Discovery Agent represents a meaningful shift in how Qlik Cloud can be used day to day. Instead of waiting for someone to open the right dashboard at the right moment, it proactively brings the data to the user. Combined with Qlik Answers for in-depth exploration, it starts to feel less like a BI tool and more like an always-on analyst keeping an eye on your KPIs for you.