Oracle Advanced Analytics is a new Oracle database option (announced today) that bundles Oracle R Enterprise and Oracle Data Mining (reviewed previously). With this release, R becomes a first class native interface for the Oracle database along with SQL and the graphic interface that ships with Oracle Data Mining.  This allows analytic modeling code to be written in 100% R with the tables and views in the Oracle database appearing as R objects directly. There is no need for modelers to write SQL – they can just write R code and manipulate the data in the database. This makes it easier for R programmers to access the database (no extracting the data to a file, no writing SQL) and the availability of data extends to Oracle OLAP Cubes that can also be accessed from the R code.

Performance is good with the approach for a number of reasons. First, with this set up, the database computing hardware is used and all the R packages are being executed on the database server. The approach further improves performance by allowing the data to be accessed without extracting or moving it. Finally all the ODM algorithms are available to Oracle R Enterprise so that R packages can use the ODM algorithms already deeply embedded and optimized for the Oracle database as well as the Oracle Exadata and Oracle Exalytics machines.

Besides improving data access and performance for R, Oracle R Enterprise also allows a piece of R code that builds a model, makes a forecast or scores a customer to be treated as a database function. Once deployed to an Oracle database function this R code can then be called by any piece of SQL (in a BI tool like OBIEE or Java code or business rules). Any arbitrary R code can be executed in this way with no constraint on inputs or outputs or the code/packages being used. This supports the increasing focus of modelers on real-time scoring by making it easy to embed R code as SQL-friendly functions that can be called to calculate a score or make a prediction right when the decision is being made.

This release builds on previous work around making Big Data available to R. The Oracle R connector for Hadoop is available in conjunction with the support for R in the Oracle Big Data appliance. These allow you to run R against both Oracle’s own Big Data appliance and against a generic Hadoop/HDFS installation with no need to convert it to MapReduce for execution. This means that a user can develop a single R script that brings in data from an Oracle database, Oracle Big Data Appliance and HDFS and have it all look like R objects to the script. These scripts can then be deployed to the Oracle Database or Oracle Big Data infrastructure for real-time scoring against this diverse set of data sources.

This will clearly be a product considered in the forthcoming Decision Management Systems Platform Report.

StatSoft was founded in 1984 and started building statistical software when it first became practical to deliver on the PC. STATISTICA is an enterprise predictive analytics platform on the Windows platform with role-based access, connections to the various data sources that companies have and support for data exploration through to deployment. The product has four main pieces:

• Windows-based analytics Workbench for analysts.
• Decision Management to combine models and rules to automate decision-making.
• Enterprise Server to support multiple users in a client/server environment.
• Enterprise Workspaces to capture the data analysis process from end to end and for managing metadata, decision-making workflow etc.

STATISTICA is a long time Windows platform user and Microsoft partner. As a result it offers a solution that is tightly coupled with Intel multi-core chips very well integrated with Windows. Everything is available as an API call in .Net making it easy to integrate into SharePoint or other Windows applications.

The components get combined into various analytic applications such as a warranty analytics solution, credit scoring, collections, cross-sell, insurance fraud detection, subrogation, price optimization, marketing mix optimization and more. These solutions can be completely automated, accessing multiple data sources, running tens or hundreds of predictive analytic models, writing results back into the database and monitoring the performance of the models.

With version 11, StatSoft released the STATISTICA Decisioning Platform that pulls together all the existing product capabilities with new rules management, integrated rules scoring, and other capabilities. The suite now includes:

• Templated data access
• Data pre-processing
• Rules management
• Modeling tools including accelerated logistic regression
• Version control
• Direct deployment

Everything is managed in an enterprise metadata repository deployed on a relational database. Workflows and other components for model creation or business rules are created, managed in the enterprise repository and deployed to a server for execution. Multiple projects and folders can be managed in the repository and permissions are layered onto these. Data access templates, analysis templates, decisioning flows and rules are all managed in this repository. Decision flows with models and rules are checked in and then used to drive reporting (integration with MS document tools), batch scoring for writing back to the production database or deployment to the STATISTICA Live Score Server for real-time decisioning using web services calls. There is a Monitoring and Alerting server for dashboards that monitor model performance and there is an integrated Document Management System for version control and approvals of models.

A decisioning flow involves several steps using the STATISTICA Enterprise Manager product. At each stage elements are retrieved from the repository based on the access defined for users and can be written back to the repository for management and reuse.

1. The first step is to retrieve data from data connection and configuration templates. Users may have access to the underlying queries or just to the data. Data from multiple data connections can be used and a wide range of ETL functions are available in the data manipulation step.
2. Data can be prepared and recoded, using Weight of Evidence for instance, and these transformations are then deployed as rules that can be versioned and reused. The rules are sequential and can assign text labels as well as transform the data. The rules are deployed to the enterprise server and can be associated with the data source. They can then be included in the defined workflow.
3. Models can then be built using various modeling techniques and embedded in the flow. A wide range of modeling techniques are supported and the workflow can create multiple models, combine or compare them etc.
4. Additional rules can be added to the workflow. The rules node contains a sequential set of rules built using an editor that has some integration with the data structures being manipulated and has a nice tree structure to allow rules to be collapsed. Temporary variables can be managed and models can be executed by the rules as necessary. Reason codes can be assigned using array handling that lets you build a set of reason codes. Rules can be reused across batch and real-time environment and multiple workflows. Rules have access to the full range of mathematical functions also.
5. The whole workflow can then be deployed to the various deployment options.

A debugger allows a set of records to run through the flow and see which transactions fired which rules. While the rules do not offer conflict detection there is some error detection (use of a variable that is not defined for instance) and some tools in the enterprise platform to see which objects refer to which other objects. Users can run multiple paths in a workflow for comparison purposes and can then use analysis tools built into the modeling environment to see what difference a change would make or which approach would be more profitable.

Besides executing the complete decisioning workflows using the STATISTICA products, all of the models can also be deployed as C, C++, PMML, Visual Basic, SAS, Java or C# Stored Procedure. The tool also supports Visual Basic scripting and this can also be used to push things into the database programmatically.

Statsoft will be one of the vendors listed in the forthcoming report on Decision Management Systems platform technologies.

I spoke to Zementis back in June of 2011 and got an update on their Universal PMML Plug-in among other things. Since then they report growing client interest with a particular focus on real-time decision-making using real-time scoring in fraud detection for instance. They have also been updating their products. ADAPA, their analytic decision deployment infrastructure, has been expanded and they have just released ADAPA 3.5. This includes support for model ensembles, segmentation, chaining and compositions – multiple models used together as a set to improve predictive power. For instance a set of regression models, one for each node in a decision tree combined with the decision tree itself allow a customer to be put into a useful segment and then scored in a way that is highly predictive for that particular segment. Support for this in ADAPA includes weighted and balanced ensemble models.

Interestingly, support for ensemble models which had previously been added to PMML 4.0 (the Predictive Model Markup Language that Zementis and others use to move predictive analytic models from modeling environment to deployment) is now being extended in PMML 4.1. PMML 4.1 is described here and also adds support for some new model types including score cards and reason codes while further improving pre- and post-processing support.

Zementis’ focus on PMML means they have a vendor-neutral approach to modeling –any model from any vendor can be deployed. Today they offer four deployment options – deployment to ADAPA for real-time decision making as a cloud, embedded or server deployment plus their Universal PMML Plug-in for in-database deployment. This last allows PMML models to be pushed into the database as a function and is supported by EMC Greenplum as well as Sybase IQ as partners. In addition a partnership with Datameer allows predictive analytic models to be deployed to Hadoop environments. Zementis hopes and expects to add more deployment options using this plug-in.

Zementis is at the forefront of the one of the main debates about PMML. A typical predictive analytic model involves a large amount of pre-processing – data is transformed from the way it is stored into more analytically meaningful attributes that are then used in the model. The challenge this creates is that moving the model definition is only half the battle as the model will need attributes that may not immediately be available in the production environment and need to be derived on the fly. PMML 4.0 added broader support for these transformations and more and more modeling tools generate this when they output PMML. Some already provide extensive support for PMML pre-processing, e.g., KNIME and IBM/SPSS.  Even if this is included in the model there is also a chance that the way these attributes should be calculated in production systems may not match how the modeler thinks about them – the IT department may have different ideas about how to derive them more efficiently. Zementis tries to accommodate both options with support for the PMML standard definitions as well as providing a tool to let you add preprocessing definitions to PMML that otherwise does not contain them.

I am a big fan of PMML and I see increasing interest in it. The increasingly rapid move to real-time scoring is driving a need to deploy models into production (rather than simply applying them in batch to existing data) and the need to model once and yet deploy in heterogeneous environments repays an investment in a common deployment language. The folks at Zementis would add that analytic modeling resources are scarce and PMML allows you to be flexible about which tool people use to build models – as long as it generates PMML the result can be deployed.

Zementis will be one of the vendors listed in the forthcoming report on Decision Management Systems platform technologies.

KNIME is an open source data analytics product based in Zurich, Switzerland that I last wrote about a couple of years ago. They have been working away on the product since then (having started development in 2004 and released their enterprise components in 2010) and have been refining their business plan at the same time. They now have 9,000 registered users/organizations and over 500,000 downloads. About 50% of these are in life sciences and 50% in BI and analytics – this represents a significant increase in non-life sciences usage. As an open source product they have 50 or more very active community developers and the company itself is 15 people (about half are at the University of Konstanz) and describes itself as “small but profitable”. Since I last spoke with them they have added a number of enterprise components and continued to develop the core open source platform.

The platform is based on Eclipse and has, in particular, been extended with lots of new node types. Like most modern analytic workbenches KNIME allows a modeler to define the sequence of steps to produce a model. The various actions that can be performed as part of the modeling process is determined by the available node types – nodes to retrieve data, process and transform that data, apply modeling algorithms (both established and more esoteric) or output results.

In addition a new decision tree viewer has been developed and data support (various databases, SAS and SPSS, and other flat files) has been improved. KNIME supports PMML generation for models including the most frequently used preprocessing operations required for a model – this was added to PMML in 4.0 and KNIME is one of if not the only vendor completely supporting this. There are lots of nodes for pre-processing and many algorithms including some of their own, some from the R Project, others from Revolution Analytics and Weka as well as lots of new visualizations.

One of KNIME’s strengths is a well defined node interface that is nicely self-contained allowing easy integration of new node types without destabilizing the whole environment. KNIME makes it easy to add nodes so that community members and partners can extend the tool with their own node types. A community site manages the community submissions and makes it easy to add any you want. Data management and execution control are also handled through well defined interfaces allowing people like Pervasive DataRush to add their own integrations.

Because KNIME is built in Eclipse they also integrate with other Eclipse projects like BIRT (an open source reporting framework). This allows these open source reporting components to be added to KNIME workflows.

Improvements and new node types are often sponsored by companies that use the product, though KNIME continues to push its own development plan. KNIME is also an integration platform for many vendors with 15 technology vendors integrating with KNIME, especially in the life sciences environment where life science companies are paying their technology vendors to integrate with KNIME to create a single environment.

The KNIME workbench is open source with registration desired but not required. This open source client version is a complete environment. Access, transformation statistics and data mining, visualization, reporting and workflow is all available in the open source with 1,000+ native and embedded nodes including text processing, social network analytics, image processing and more. Companies can add support to the open source project or adopt the server or enterprise components that are proprietary and licensed.

KNIME Team Space allows a small group to have a shared repository. KNIME Server adds the full server-based capabilities including remote and scheduled execution of modeling flows, a workflow repository, shared data in a managed repository, shared metanodes or component workflows for reuse, and web browser access. The browser interface exposes workflows to users who can enter the parameters defined for the workflow and get a result displayed back in the browser. An API also allows “headless” execution of workflows. With the full suite companies can write new nodes and wrap legacy software, then allow power users to develop templates and shared workflows, provide this to scientists or others who adapt and use these workflows as well as managers who just run pre-configured ones.

The company is shifting from selling directly to life science and a small number of other companies to an approach based on partner channels with domain expertise. Target markets are obviously life science as well as data mining/predictive analytics (customer intelligence especially). KNIME’s value proposition is in its openness as well as the ability to reduce the cost for existing heavy users of analytic workbenches while also targeting smaller companies that can’t afford the large vendors.

Besides the partnerships with Pervasive and BIRT noted, KNIME has a strong partnership with Zementis for PMML execution in database and on servers and with Dymatrix, a Europe-based partner offering model monitoring and automated updates.

KNIME will be one of the vendors listed in the forthcoming report on Decision Management Systems platform technologies.

Quiterian is a Spanish company with offices in the US, Mexico and Europe. Quiterian Analytics aims to be complementary to traditional tools for reporting by helping companies get more value from their data sooner. In particular they aim to help companies anticipate the future by providing simple to use predictive analytics and by empowering users while reducing IT costs. They see the BI market in three broad segments:

• Dashboards and reports – traditional BI tools as well as some of the newer, easier to use visualization tools – with little or no focus on advanced analytics
• Data mining tools focused on professional data miners and statisticians, typically not terribly visual.
• Visual Data Mining tools that site between the two aimed at analysts and business users with no dependency on IT or data miners. This is where Quiterian is targeted

They offer fast and comprehensive data analysis and visualization, self-service for non-technical users and aim to be very agile with no need to pre-create cubes etc. Specifically Quiterian Analytics is a web-based SaaS tool that includes functions for

• Its own analytical database that is a columnar/in memory hybrid
• Data acquisition, exploration and engineering
• Advanced analytics
• Distributing and sharing findings
• Workflow

Users can identify various data sources and load them up using Quiterian’s ETL. Drag and drop exploration tools allow users to profile the data, examine the underlying records etc. Charts of distributions, various statistical measures and frequency distribution can be displayed easily. Subsets can be selected from the graphs and everything recalculated based on those subsets.

Pivot tables, venn diagrams, bubble charts and more can be used to see what is going on in the data. For instance the customers who don’t have a credit card but do have a current account and don’t have a credit card from another company. These analyses can be saved, exported and shared in a variety of ways.

Expressions, calculated values, can be defined and used in analysis. Users can also use more advanced tools such as aggregates across rows, tools to manage quartiles/deciles, or tools to apply Pareto distributions to see what percentage of profit comes from a certain percentage of customers for instance. Profiling tools can be used to assess correlation, for instance by identifying which attributes are strongly correlated with profitability. A profile can then be applied to a group based on a previous query or analysis. New selections can be based on these profile attributes and pivot tables and other kinds of analysis can build on these selections.

In addition the tool supports decision trees, clustering and forecasting for more advanced analytics. The decision tree data mining algorithm has some nice usability features with color coding, accuracy measures etc. This can then be saved as an algorithm and be applied to any set of records selected in the tool using any of the other elements.

Finally the workflow component allows tasks within the tool to be scheduled. You can schedule actions to take when a new model is created, as part of a scheduled campaign, when specific technical changes are made or just at a particular time. These tasks can apply models or algorithms, send emails etc. Quiterian is working on adding some specific action types to expand this e.g. invoking actions in salesforce.com or email marketing tools as well as twitter, SAP, Omniture and others. As the range of actions expands, the ability to embed these analyses into decisions will likewise expand. A PMML import/export capability is also in the works.

Talend has been around for about 6 years and the original focus was on “democratizing” data integration – making it cheaper, easier, quicker and less maintenance-heavy. They originally wanted to build an open source alternative for data integration. In particular they wanted to make sure that there was a product that worked for smaller companies and smaller projects, not just for large data warehouse efforts.

Talend has 400 employees in 8 countries and 2,500 paying customers for their Enterprise product. Talend uses an “open core” philosophy where the core product is open source and the enterprise version wraps around this as a paid product. They have expanded from pure data integration into a broader platform with data quality and MDM and a year ago they acquired an open source ESB vendor and earlier this year released a Talend branded version of this ESB.

Talend 5 is a “holistic integration platform” with a newly announced OEM of BonitaSoft (they were already using the workflow capabilities of BonitaSoft in their MDM product). Talend 5 now embeds the complete Boniasoft product. This gives them data integration (historical core), process integration (BonitaSoft) and application integration (ESB)

There are 5 elements to the core Talend platform

• Common design environment across all three elements, including the BonitaSoft elements being OEMed
• Shared project repository for artifacts and metadata
• Single deployment approach
• Single runtime environment
• Single monitoring and management platform

A single platform really helps with projects that combine multiple elements (data integration and BPM say) – Talend try to make it easy for organizations to adopt additional elements by reducing the learning curve through common tooling etc.  Nevertheless Talend remains committed to providing best of breed elements in each case. The product supports the increasing array of cloud and hybrid-cloud deployments with particularly strong support for hybrid clouds. They support “Big Data” quality and integration tools that run inside Hadoop and allow companies to handle this data the same way they would handle data stored elsewhere. They also embed the JBoss Rules engine and JBoss Guvnor for rule management. This obviously allows for both simple business rules on the integration and data quality side and more complex embedded decision-making.

Good decision-making relies on data being delivered to the point of decision. An approach to data integration that focuses on the use of data for a specific solution while allowing a broad-based, coherent integration architecture to be developed incrementally feels like the right approach to data integration in a decision management context.

2011 has been a great year for market awareness of Decision Management as an approach and of the value of Decision Management Systems. Product, partnership, acquisition and funding announcements have enhanced the available technology. As vendors continue to improve and enhance their product offerings to fully support Decision Management this is only going to reinforce and further grow the market.

There is a wide range of technology available for building Decision Management Systems. Business Rules Management Systems, Predictive Analytic Workbenches and Optimization technologies can be used alone or in combination to build custom Decision Services. In-database analytics and other analytic infrastructure can be used to maximize the effectiveness of predictive analytics.

In each Decision Management Systems platform product category, products will be reviewed and key elements for successful Decision Management Systems (such as for business user analytic modeling, in-context rule management, decision simulation etc) identified and highlighted with “badges”. Customer case studies and use cases will be used throughout. An index of vendors and product categories will be provided. The list of vendors identified to date are listed below (if you have suggestions for additional vendors, please drop me a line at info@decisionmanagementsolutions.com).

The report will be available free to download as a white paper and as a web page at decisionmanagementsolutions.com/decision-management-technology where we currently have links to our existing First Look product reviews.

Obviously one of the challenges with a report like this is deciding who to include. With that in mind we have identified a few criteria:

1. While there are many innovative companies developing complete solutions – pre-configured Decision Management Systems – we will be focusing in this report on tools for building custom systems. To be included it must be possible to build a custom Decision Management System with the product.
2. The technologies must be productized, released, for sale, and must have at least one production customer who can be contacted.
3. The product should be sold generally, not only to companies that have also bought a pre-configured application – it must have customers and a market presence as a genuine platform technology.
4. The business rules, analytic or optimization technology involved cannot only be an OEMed product that is included separately in the report.

There is no fee for vendors to participate and everyone will be covered, sponsor or not.

We have begun work on this report and will be releasing it over the first two quarters of 2012. The first installment of this report will include the business case for Decision Management Systems, the overall technology architecture and a list of product categories and vendors. Over the following months we will publish First Looks on all the technologies involved and several updates to the report:

• Initial report early Q1 2012
• Report update with “badges” late Q1 2012
• Report update with customer stories and use cases early Q2 2012
• First complete report late Q2 2012
• Report update for new/revised products Q4 2012

We will announce each new release and all the related activities in our newsletter, which you can subscribe to here.

Candidate Vendors

• 11Ants
• Angoss
• Attensity
• Be Informed
• Bosch Software Innovations
• Clario Analytics
• Experian
• FICO
• FuzzyLogix
• GDS Link/Modellica
• Gurobi
• IBM
• IDIOM
• Infocentricity
• InRule
• Jboss
• KNIME
• KXEN
• MathWorks
• Microsoft
• OpenRules
• Oracle
• Pega/Chordiant
• Predixion
• Progress/Corticon
• Revolution/R
• Salford Systems
• SAP
• Sapiens
• SAS
• Sparkling Logic
• Statsoft
• Teradata
• Tibco
• Transunion
• usoft
• Yottamine
• Zementis
• Zoot

Don’t forget, if you have suggestions or comments on the vendor list please email us at info@decisionmanagementsolutions.com.

Starview Inc. have just launched their new platform – Business Analytics and Optimization Platform (BAOP). Growing out of a long standing consulting practice, the company sees an opportunity for a platform aimed at the continuous optimization of decision-making. They are focused on Telco, Finance, Energy and Manufacturing (especially Semiconductor) and they have a vertically-oriented strategy, working with partners that have domain expertise to deploy vertically oriented solutions.

Starview has been developing technologies to help companies deal with “Big Data” such as time series data out of factories and management systems. These companies are naturally moving to analytics on this data and ultimately to automated decision-making given the velocity of data and the need for rapid responses. Starview felt that there was no platform for this kind of system out there and that piecing it together from various platform components was too hard. They have been developing an “active model” platform that allows them to define actors that represent the various elements of the system. These actors understand state and can respond to the data as it flows in and around the system. The platform allows various algorithms to be applied across the actors to optimize decisions. See a video here.

For instance, continuous manufacturing. Traditional factory optimization is very batch oriented and Starview wanted to do this continuously – generating a “ best next action” for the factory or complex of factories. The same approach can be applied in Telco to ensure that high value customers get prioritized as they move around – optimizing their use of the nearby towers relative to other (lower value) customers. In financial services Complex Event Processing has been big in algorithmic trading and Starview is helping companies look at their portfolio in the context of everything that’s going on to make more optimal trading decisions that use a broader view of the situation.

The core of the solution is a model, composed of actors that understand state (up to millions of them) across which decision logic is run to make decisions. Simulation is important in advance and there is also some “in line” simulation for impact analysis when making changes. For instance in a system managing a smart grid you need to understand the consequence of more people plugging in Electric Vehicles. How to manage the grid in these circumstances to make sure that charging the EVs does not overload the grid and disrupt other uses of the grid. The simulation could allow for a given number of EVs, particular growth rates etc.  View Starview’s Smart Grid demo here.

A typical buyer  is a “business user” – someone who understands how to run a factory or a network not someone who understands coding. The system allows the definition of management rules and analytics to make sure that it behaves appropriately. In the Smart Grid example, for instance, you can define logic to ensure the grid load stays in bounds and spreads the charging of EVs out appropriately.

The development environment is based in Eclipse and allows the definition of the model using various diagrams. The core diagram is a data flow model that identifies the various actors whether systems or data sources or decision-making components. Each has a well defined set of interfaces for data transfer and the structure is hierarchical so you can have the same kind of actors defined at different levels of granularity. Large numbers of adaptors are available to pull data in from other systems as you would expect and these support real-time data feeds.

The system is very message based and chains together actors into complex webs. Within each component there is a bunch of code. You can code as much complexity as you need but it is still being coded. It’s frankly a fairly geeky interface, typical Eclipse, but there is a rules-based language that is declarative and allows large numbers of rules to be specified in any order. These rules can be changed in running systems by specifying a name for the rule and then building a UI to update the rule from within another application. There is a standard management console that allows attributes to be changed but most customers like to build a custom interface and use an API. When rules are changed the component is paused while the update happens and buffers the data flowing in until the change is complete.

I got an update from InfoCentricity recently. InfoCentricity are a software company focused on delivering a web-based, advanced predictive analytics workbench (Xeno). They were founded back in 2000 have been releasing various components of their Xeno platform since then as well as a first application based on this platform (Campaign Analyzer). Xeno4 is the new architecture and is starting to roll out with a new Decision Trees product (Trees4) in 2012 (the rest of the platform to follow in 2012/2013). Their initial focus has been financial services and retail marketing as well as the data bureaus serving those markets. Most of their customers access the software as SaaS and also work with InfoCentricity for analytic consulting support when building advanced predictive analytic models.

InfoCentricity is focused on helping clients extract the maximum value from their data by deriving actionable insights. They see themselves as transformative not disruptive – adding value to existing analytic infrastructure by being collaborative, data store agnostic and fast/flexible as well as by providing knowledge-transfer through their analytic consulting. InfoCentricity sells directly in the US (and through partners internationally) with a development team based in Novato, California.

Xeno Pro modules are Pro Scorecard, Pro Trees and Pro Clusters. These sit on a base layer supporting variable generation, analysis and reporting and data import/export.

• Pro Scorecard supports automated performance inference and all the various predictive scorecard bells and whistles. The new version supports continuous predictors allowing, for instance, compound scores based on inputs that are also scores, to be developed more effectively. This also supports multiple performance objectives – say a primary objective of managing risk and a secondary one of revenue. This product is squarely aimed at those trying to build the most sophisticated predictive scorecards possible.
• Pro Trees supports classification and regression, segmentation and decision tree/strategy development and provides a highly interactive tree builder environment – data driven but analyst guided. This is in limited release at present and the full release is expected soon on the new platform.
• Pro Clusters supports exploration and segmentation development. InfoCentricity’s clustering implementation handles many of the data challenges inherent with traditional k-means clustering such as discrete variables, missing data, and variables of different scale.  Typical workflow is supported with tabular and graphical cluster profiles for ease of evaluation.

Users are large financial services companies (some of which have 100+ modelers using the product), credit data providers and direct marketing vendors. These are overwhelmingly also users of other sophisticated analytic tools and use InfoCentricity to boost the performance of the models they build.

The Xeno4 architecture is based on EXTJS for UI, Zend for the business layer, SQL DB and a proprietary vertically partitioned database and C/C++ for the core analytic engines. The whole architecture remains browser based but is moving to Web 2.0 with more drag and drop and an increasingly visual user experience. They are working to move all their existing products to this architecture while continuing to develop the engines and expect to be done in 2013. The product mostly runs as SaaS from a pair of co-location facilities but can also be run onsite.

Netuitive provides predictive analytics for IT. Based in Reston VA and founded in 2002 they have over 50large enterprise customers and 300 more through OEMs. Their solution is designed to prevent degradations and outages to critical applications and services by providing an intelligence layer on top of existing monitoring systems. Companies use the software for critical systems where either there is revenue dependent on the system being available or where keeping the system up is critical for a company’s reputation (like online banking for a bank for instance).

These critical systems involve lots of pieces. Each is monitored separately and these monitoring systems and key performance indicators are not connected. In addition there is just a lot of data. Netuitive has a patented behavior learning engine that consumes outputs from the applications, physical systems and virtual infrastructure and uses predictive analytic approaches to create an application health score. It can consume data from any source, any monitoring system, and correlate it

These health scores are presented on dashboards or in alerting systems. They can also generate “trusted alarms” that show that things are heading in the wrong direction or getting out of bounds relative to what would be expected. The scores come with diagnostic information and alarms are very focused, targeted at specific people who can address the problem and provided with root cause analysis as part of the alarm. These trusted triggers could drive system behavior too, reconfiguring to move a VM for instance though most users are not using the software in this way yet.

Netuitive claims to have algorithms and approaches that allow them to do this kind of thing without lots of processing power. They can plug in lots of systems and the software will learn about the performance dependencies in those systems. In addition the user of the system can connect the pieces to show which bits contribute to which business capability.

An interesting way of applying predictive analytics to some of IT’s own problems, this also illustrates the power of self-learning or machine-learning in situations with a lot of complexity.