Understanding Data Insights

AAI can extract, aggregate, and abstract huge amounts of workload automation data to give you precise, reliable views into your workload executions and scheduler activities, even in real-time. You can create these data extractions, called data insights, based on provided templates that present the data in an easy-to-consume format that is logical and understandable even to people in non-technical roles. You can then either use the results for yourself or make them available to others for on-screen viewing, as reports, or as dashboard widgets.

Because AAI can as easily pull real-time execution data, these data insights are a useful tool for operators who monitor daily executions, as well as their supervisors. The historical data insights are the most important tool for reporting both internally and externally, and for all levels of data analysis for triage or performance improvement of jobstream executions, as well as for compliance audits.

Tabular and Graphical Information

Each data insight provides a data table that can contain all the details related to the filtered data, as well as a graphic.

You can configure the presentation of the data in the table, to change sort orders and hierarchies and to hide some columns while rearranging the rest so that you have the view that is most useful to you. The table data is included in both PDF and CSV output that you can produce from the data insight. The data in the output file is filtered and configured just as you see it when the output is triggered.

You can choose how the data is grouped in the graphic and then click on the data segments to quick filter the table to include only the rows that belong in that segment. The graphic is also included in PDF output that you can print from the data insight.

For more information, see Viewing and Using a Data Insight.

Types: The Templates for Data Insights

You create data insights based on templates for various types of data extractions. When you define a new data insight, you start by selecting its type. Each type is a template of select data extraction that supports a specific purpose. These can be direct reads from the AAI database or complex interpretations and calculations based on values from the database. Having the data insight type already pull and prepare relevant data from the rich AAI database simplifies and error-proofs a process that would otherwise require you to create complex queries and data manipulations.  

When you add a new data insight for a certain type, the predefined data values are the data columns for the table in the data insight. You can filter to include only a subset of the data rows by specifying the values for the data columns that you want to see. You can also choose to hide or rearrange some of the data columns to see only what is relevant to your inquiry and in a structure that surfaces the clues and indicators that you are looking for. For more information, see Viewing and Using a Data Insight.

Historical and Real-Time Data Insights

The configuration of each type of data insight includes a time range specification, so you can define the time span that is useful for your purposes. There are two kinds of time range specifications that support the type of data that you extract:

• An absolute time range for long-term, historical analysis

  When you define an absolute time range, you define specific start and end dates in the past. This extracts historical jobstream execution data from AAI's historical database and aggregates them as required. Data insights for an absolute time range can include jobstream execution events and statistics to help you discover trends and patterns that can give you clues to how to improve your jobstream processes or scheduler workloads.

  Historical data extracts capture jobstream execution events and statistics for specific time ranges can also be used for reporting of all kinds.

• A relative time range to include real-time analysis

  When you define a relative time range, you specify how far back from now until now you want to include in the data insight. For example, you might want to only extract data from the last 12 hours, the last 30 days, or 90 days, and so on. These data insights include both historical and real-time jobstream executions. These provide the input for real-time analysis to view the most recent trends, moving averages and currently running hotspots where jobstreams have failed or are at risk of missing their SLAs.

The default time range for a data insight is the last 30 days, which includes real-time data.

Dynamic Data Manipulation

A useful feature of every data insight is that you can change the filters, sorts, and data column configurations dynamically while viewing the data insight.

This allows anyone who can view a data insight to manipulate which data is included and how it is presented to gain different perspectives on the dataset. By exploring the real-time data dynamically, you can reveal the patterns and outliers that tell the story of how your workload schedulers and jobstreams are performing. By playing with the data, you discover clues to where snags and problems that are affecting successful SLA compliance lie. You also confirm where processes are running smoothly or can be further optimized.

When you finish your exploration, you can close the data insight to reset it to the original definition, or choose the option to save a copy of the data insight to create a new one with the current filter and configuration.

Anywhere along the way, you can capture the current view as a printed output or in a CSV file to use for your own comparisons, or to share with other interested parties.

For more information, see Viewing and Using a Data Insight.

The Source for Dashboards

Data insights are the source of the widgets that you can add to your personal dashboard. A dashboard widget is a smaller presentation of a full data insight, a window to the data insight with interactive functions that allow you to dynamically play with the filters, sort orders, column configurations, and quick filters to bring into view the data that you need. You can add any data insight that you can view or edit as a widget to a dashboard.

By adding different widgets to your dashboard and organizing them on dashboard tab pages, you bring different views of complex execution and scheduler workload data side by side onto one canvas. This can be helpful to make comparisons among them or to just have the data views that are most important to you readily available at a single glance.

For more information, see Dashboards (Web Interface).

Scheduling

For regular reporting or for archiving of jobstream execution and workload scheduling data, you can schedule any data insight to be run automatically on a schedule and emailed to appropriate people. With scheduled output, you can routinely capture a snapshot of various facets of your workloads and provide consistent reporting to both company-internal stakeholders as well as customers, keeping everyone up-to-date with the verified data that they need.

Also, because you can schedule CSV files as the output, you can regularly provide data in a universal format that can be consumed by other reporting or analysis products that your company uses or requires.

For more information, see Scheduling Data Insights.

Sharing

All data insights can be shared with other users to view in real time. Share options include making it available to all AAI users to view, selecting specific users with either just view rights or with edit rights.

Sharing a data insight is the only way to give another user access to your data insight in real time. Everyone you share the data insight with can see the filtered data and configuration as they are specified in the data insight definition. More interestingly, while viewing the data insight, the sharing users can play with the filters and column configurations to dynamically manipulate the data on their view, just as you can. This allows them to focus on different slices of data and relationships to surface data that is significant to their goals.

Since they can print at PDF of the view they have created or download a CSV file with the data, they can capture their own data snapshots as they explore. Furthermore, they can add shared data insights as widgets to their own dashboards.

For more information, see Sharing Data Insights.

Printing and Downloading

You can run a data insight to produce an output file in the format you choose, such as PDF or CSV. The data in the output is based on the filters and column configurations at the time the output is triggered. This allows you to capture the data as you are viewing it and with any dynamic changes you made to the filters and configuration. You can keep these outputs for your own use or send them to other AAI users, colleagues, internal stakeholders, and customers.

A PDF file contains the data table and the graphic from the data insight. A CSV file contains the data in a universal format that you can feed into other reporting or analysis tools or processes that your company uses or requires.

Anyone who can view a data insight can print a PDF file or download a CSV file at any time. These outputs can also be scheduled to be run at regular intervals and emailed to appropriate stakeholders.

For more information, see Actions for Downloading, Printing, and Emailing Data Insights, Printing Settings for Data Insights, and Scheduling Data Insights.