From real time to deal time

Business intelligence and analytics apps from Iteration and Spotfire empower all kinds of users to make smarter decisions

The emerging field of BA (business analytics) solutions extends the capabilities of BI tools with sophisticated mathematical and statistical routines and improves the ability to interact with data.

There are a wide range of tools and systems in the BA and BI galaxy. Two such noteworthy solutions, Iteration Real-time Reporting Suite and DecisionSite, deliver on innovative goals, enhancing the way organizations deal with the tidal wave of data today’s enterprise typically generates.

The Iteration Real-time Reporting Suite exemplifies how pure BI is expanding. It does a number of useful things by adding real-time turnaround to data-warehouse reporting. Iteration’s impressive interface allows those who haven’t mastered reporting technology to craft deliverables though a familiar PowerPoint-style interface, which they’re likely to already know.

Because it integrates well with other data sources, one could merge it with BA tools for more refined future data exploration.

Spotfire’s DecisionSite, a BA system, puts advanced mathematical analysis into the hands of all kinds of users in a collaborative environment that supports close teamwork. It features “guides,” essentially wizards, that allow analysts to build sets of queries, filters, and highlights so that domain experts who lack statistical knowledge can receive and then explore organized data sets.

Iteration Real-Time Reporting Suite

The Iteration suite is a real-time business management tool. Unlike traditional data warehouse products that rely on a batch-oriented ETL (extract, transform, and load) cycle, Iteration processes and presents business data as a constantly updating stream of information. By making real-time feedback consoles available to data-rich enterprises, it aids in making quicker, better-informed decisions.

The installation of the Iteration suite is fairly straightforward, after the prerequisites have been installed. There are surprisingly few configuration parameters or installation choices. The product relies on the integrated security feature in Windows for single sign-on support.

At the headwaters of the data stream is the Enterprise Link server, based on the Sagent Data Flow Server. Enterprise Link is responsible for capturing transaction messages from back-office systems, EAI systems, existing data stores, data warehouses, OLTP (Online Transaction Processing) systems, and the like. The Data Flow Server transforms messages on the fly using the Sagent data-flow language and then sends them on to the ADC (Active Data Cache).

The ADC is the heart of the Iteration system. Once there, the system uses messages to update or add to the existing information. The system keeps the data entirely in memory. Performance figures bear out the wisdom in this design: The ADC can process up to 11,000 records per second. On the downside, this design places a heavy burden on the hardware. Iteration recommends a server with at least four processors and 8GB of memory to get started.

To understand the downstream architecture, it is critical to visualize the way reports and alerts work in Iteration. An Iteration report is like an information dashboard that updates in real time as the information streaming into the system changes. So for example, you can create a report that has as one of its fields a moving one-hour average of call center volume; you can then see the numbers or visualization gauges changing second by second as the volume goes up and down.

The Event Engine sits downstream from the ADC and manages triggers and events on the data in the ADC. The Event Engine routes relevant data to the Reports Engine or to the Messaging Engine for alerts.

Alerts are an important feature of this events-based system. IT can configure the Messaging Engine to send a page, an e-mail, or an instant message according to a set of pre-established rules. The system monitors alert responses and escalates the alert up a predefined path according to a set schedule. Alerts can be time based, or scenario based (for example, whenever sales dip below a certain amount).

The Iteration system has three main configuration points: data extraction and transformation in Enterprise Link, reports in the Reports engine, and alerts in the Messaging Engine. All are managed from a single configuration console implemented as a light client using VML (Vector Markup Language) in Internet Explorer. The VML-based interface is fluid and feels more like a thick-client GUI.

The administrator configures data extraction and transformation using a graphical data-flow language. The administrator selects a data source and then directs the flow of information through various filters, splitters, programming agents, and data sinks. You can drag and drop various pieces onto the programming palette and connect them up to indicate how data flows through the system.

Reports are meant to be created by business users Consequently, the tool is easy to use, borrowing the PowerPoint model of designing slides. Each report contains views in a variety of configurations, and each view has a view type, which can also be configured. Using the report tools, anyone in the organization can create real-time dashboards containing any of the information they’re authorized to view. Reports can be shared, including via e-mail.

DecisionSite7.2

DecisionSite 7.2 is an interactive, visual exploration BA tool with templates for both horizontal and vertical applications. The client supports the application of advanced and intermediate statistical routines, and it presents sorted tables simply.

Many features set DecisionSite apart. The breakthrough lies in the collaborative design that divides the logical and productive applied analysis cycle into a three-tier user architecture, a model that unleashes the power of many knowledgeable end-users who would have been isolated from the process. This trisection provides a separate, user-appropriate interface for each of three kinds of user: analyst, domain expert, and end-user.

The first, smallest tier is the group of analysts who can master the statistical client itself, the DecisionSite Browser, where they plug into a predefined information library, drawing information from various local and enterprisewide data systems.

After analysts loads the data sets, they can start looking at and manipulating various kinds of graphs, plots and charts, check relationships of two or more fields with selections and filters, and see what changes the shape or significance of the results.

The analysis tools took some getting used to, and some controls didn’t work as we expected. But the online help, in combination with a couple of hours using the system, was more than adequate to get comfortable.

The client proved responsive in what-if exploration. Graphical and statistical displays update in a flash, even when the data selected is radically altered.

Once analysts complete the work of exploration, hypothesis, and confirmation, they use the Analysis Builder to create a guide, a wizard that domain experts can use to examine data and hypotheses without needing expertise in manipulating the browser.

Analysts build guides using a sophisticated macro builder, much like a tool for building a LotusScript routine or a FileMaker script. The analyst selects the Analysis Builder then runs the desired browser exploration sequentially, adding explanations or other documentation and directions within the script, as well as customizable names for each step.

Builders can make scripts that take input from a guide user, which makes exploration more interactive and allows downstream users to add value to the process. The scripting function allows a builder to put multiple steps under one title, so the process appears simpler to the guide user. This makes for clear directions and an nonintimidating interface that can share insights that the Analysis Builder user notices.

Building and executing complex guide scripts was fast and straightforward. Debugging and repairing the mistakes we made was surprisingly easy. The intrinsic tools for adding fields or doing operations on one or more fields are obvious.

After scripts are debugged, an analyst distributes and manages guides from the DecisionSite administrator tool.

Spotfire has designed the workflow so that analysis builders and guide users develop, test, and refine strategies for excellent collaboration. To ensure everyone is on the same page, Spotfire uses “posters,” Web presentations that capture the underlying data along with findings, an analysis session, and all the team’s communications and annotations.

Because posters are simply presentations that don’t require specialized knowledge of the system, analysts can use them to expand the decision-making team by bringing in expertise that wouldn’t otherwise be available to the team.

The system’s architecture includes a poster library to hold and organize these findings, which could inspire collaboration and empower teams with information.

Iteration can be extremely useful to IT groups trying to add value to the enterprise by monitoring business activities in real time. Real-time feedback consoles can aid in making more informed, quicker decisions.

For shops that want to put data to maximum use, DecisionSite is worth a close look.

InfoWorld Scorecard
Reporting (20.0%)
Interoperability (20.0%)
Scalability (10.0%)
Suitability (20.0%)
Value (10.0%)
Ease of use (20.0%)
Overall Score (100%)
Iteration Real-time Reporting Suite 9.0 7.0 7.0 8.0 8.0 9.0 8.1
Spotfire DecisionSite 7.2 9.0 8.0 8.0 9.0 6.0 7.0 8.0
From CIO: 8 Free Online Courses to Grow Your Tech Skills
Join the discussion
Be the first to comment on this article. Our Commenting Policies