Pitney Bowes Spectrum Technology Platform

The Spectrum Technology platform is the foundation of the Pitney Bowes Customer Information Management solution. It is a modular platform, delivering comprehensive data management and governance capabilities that includes almost all the sort of functions that one would expect from such a platform: data profiling, data matching, data cleansing, a data stewardship module, data integration (including ETL, data federation and change data capture) and master data management (see http://www.pitneybowes.com/us/customer-information-management.html). The only thing that one might argue that it is lacking is a business glossary. On the other hand, it has a number of capabilities that one would not expect, especially around relationship discovery, entity resolution, advanced analytics and Location Intelligence (see http://www.pitneybowes.com/us/location-intelligence.html), which are particular strengths of Pitney Bowes’ offerings. Further, there is a Global Sentry module, which identifies people on government watch lists and, in the United States, there is an Enterprise Tax module that determines the tax jurisdictions for an address for accurate sales/use tax, property tax, payroll tax and insurance premium tax calculations. In addition, Spectrum integrates with Pitney Bowes’ Customer Engagement Solutions (see http://www.pitneybowes.com/us/customer-engagement-marketing.html) to provide a combination that supports a single, contextual view of the customer along with more effective engagement via EngageOne Video, which provides an interactive and personalised video solution.

It is also noteworthy that the Spectrum Technology Platform includes analytics capabilities which are based on the Portrait Software solution. Analytics is modularised and includes visualisation capabilities, predictive analytics, real-time customer scoring, and an interaction optimiser and uplift capability, which is designed to help marketers focus on the audience that is most likely to accept a particular offer or which will produce the greatest return. Alternatively, it is possible to embed predictive models built on third party products into Spectrum data flows through the use of an industry standard interface called PMML (Predictive Modelling Mark-up Language).

Pitney Bowes uses a direct sales model for coverage across its key geographies in the Americas, EMEA and APAC with an extensive partner network in other locations. Its solutions are applicable across industries, but the sales teams and solutions are vertically aligned with a use-case (or solution) focused approach that is relevant to that industry and personas within that industry. The company’s go-to-market approach is also augmented through strategic alliances that are global in nature, but operationalised at a regional level. In recent years, Pitney Bowes has also addressed a broader audience of users through its SaaS offerings. These include APIs as part of Spectrum On-Demand hosted on AWS, apps on marketplaces like Office 365 Excel and NetSuite, and Amazon Machine Images on the Amazon B2B Marketplace that are available for use on an hourly, monthly or yearly subscription basis.

Spectrum Technology Platform customers are spread across many industry sectors, with most of the client base represented within the mid-to-large enterprise segment and the Global 500. These include users of the company’s legacy data quality and integration products as well as those that have adopted or are new users of the Spectrum data platform.

The Spectrum Technology Platform uses a services-based approach with an orchestration layer that links and integrates the various data management and analytics functions together to deliver composite business services focused on business outcomes.

Perhaps most interesting from a technical perspective is that both the MDM repository (technically, the Data Hub module) and the system repository (which may be distributed) are based on graph databases. Or, more accurately, MDM is based on the market leading graph database (Neo4J) and the system repository on what can best be described as a hybrid document-graph databases (OrientDB). These have significant advantages when it comes to exploring master data and metadata respectively, because they allow you to see and understand relationships more easily and quickly. In the case of metadata this makes processes such as data lineage much easier while in the case of MDM there are specific abilities (Relationship Analysis) built into the product that allow you to model and explore customers and their structures and their relationships to products, locations and other details that would not normally be easy to achieve, let alone be operationalised for real-time systems. The use of Neo4J, a fully ACID property graph database, enables users of Pitney Bowes’ MDM capabilities to craft and deploy a contextualised view of their most important data assets and their relationships inside operational systems, and not just for the purpose of analytics.

Neo4J’s claim to fame is its ability to cater to high volume, real-time transactional queries where many relational databases fall short, given the cost of complex joins to understand highly connected data structures. In an increasingly connected world where everything from devices, to computers to wearables to assets to social and professional networks intersect, organisations can struggle to understand this increasing depth and complexity in data relationships. And without that understanding it is equally difficult to derive new insights from all of this information, without a strategy to evolve and enhance existing MDM infrastructure. If that isn’t reason enough to drive adoption of this technology, for those organisations choosing to walk before they run, this technology allows organisations to take a much more agile approach to MDM.  

The company is also building out industry specific applications that leverage the Spectrum platform, there is a common set of APIs across all functions, and both Hadoop and cloud connectivity and deployment options are provided, including certifications for Cloudera, Hortonworks, IBM Apache and Map R.

Pitney Bowes provides implementation services, support, enhancement and upgrade services, managed services offerings, advisory services, business consulting and solution add-ons, as well as complementary products that span both data services and predictive analytics. It also relies heavily on a partner network to deliver its services in regions such as Latin America, Middle East, Eastern Europe and Japan.