Essentially, any data virtualisation platform does four things. Firstly, it virtualises your data (making all relevant data sources - structured or unstructured - appear as if they were in one place so that you can access that data as such). Secondly it abstracts your data (presents the data to relevant applications in a consistent fashion regardless of any native structure and syntax). Thirdly, it federates your data (allows you to pull data together, from diverse, heterogeneous sources and present that in a holistic manner, while maintaining appropriate security measures and quality of service). Lastly it presents your data in a preferred use-specific format to the front-end application either through relational views (via SQL) or by means of data (web) services of various types.
There are a variety of reasons for deploying such technology, which include but are not limited to: agile business intelligence; enabling federated query environments where there are multiple data warehouses or data marts in use (logical data warehouse); support for master data management (MDM) environments, where the MDM hub or registry needs to be kept synchronised with source systems; integrating enterprise systems such as ERP and CRM with external data for "single-view" applications; support for a service oriented architecture and agile application development by providing reusable data services; integration and access within the Internet of Things; and for cloud and SaaS-based integration. Data virtualisation is also useful for prototyping warehouses, MDM and other applications.
Data virtualisation is not specific to any one market segment and it is applicable across vertical boundaries. As a result Denodo has no particular industry focus. More generally, the company uses both direct and indirect marketing channels. In the latter case it has both resellers and service partners on the one hand and technology partners with whom it co-sells on the other. Perhaps the most notable of the latter is IBM but the company also partners with Tableau and MicroStrategy amongst others.
In order to encourage usage and take-up of its platform the company has recently introduced a free-to-download (limited) version of the Denodo Platform called Denodo Express.
Denodo has some several hundred customers. As data virtualisation is applicable across all industry sectors it is not surprising that the company's clients range across a diverse set of industries. They include Caterpillar, Telefonica, Vodafone, several pharmaceutical companies, Nationwide Insurance, government agencies, AAA and BNSF Railways, amongst others.
In terms of connectivity it would be difficult to find a data source that Denodo could not connect to. It supports ODBC, JDBC, MDX, XML, JSON and so forth, has widespread NoSQL connectivity, (including graph databases) provides access to legacy hierarchical databases, and so on. If you can find something that isn't currently supported then there is an Eclipse-based software development kit available. On the output side you can build (relational) views or use web services (often using a RESTful API) to expose query results.
For modelling Denodo offers a choice between a bottom-up and a top-down approach. Using the former approach, you define your connections, create normalised views, integrate them, and then build an appropriate data model for data consumers. Using the top-down approach you start with business views that represent a contract with the consuming applications or users and then derive your data model, leaving physical definitions for later. This can be particularly useful in prototyping environments, parallel development, migration of sources, and contract-first web services. Note that you can import data models built using external data modelling tools, from industry models or other sources or you can do this using Denodo's own graphical user interface, which is wizard-driven.
Perhaps the biggest issue with respect to data virtualisation is performance. Denodo provides an optimiser (with support for query delegation, automatic query rewriting, join optimisations, caching and data movement, cost and source constraint based query plans, asynchronous multithreaded processing and server throttling mechanisms), support for caching (both full and partial caching), a scheduler, push-down optimisation (where sub-queries run locally), and workload management. Performance tests, executed in conjunction with IBM Labs, have shown that the Denodo Platform has minimal impact on query performance and, in some cases, can actually improve overall system performance through its advanced optimisation and intelligent caching strategies
Finally, fine-grained security provisions are included in the product and the security model is extensible and integrates with existing security mechanisms. This has been validated by the U.S. National Nuclear Security Administration, which integrated the Denodo Platform with its internal security mechanisms to protect classified nuclear weapons data.
Denodo provides training for architects, developers and administrators through both public course and private training. It also provides free training videos on its web site. Other than training there are the professional services and support offerings one might expect. There is also a dedicated partner portal as well as a community site. This last includes documents, Q&A, videos, tutorials and a knowledge base that provides a series of technical articles.