This section contains resources related to Cortex.

CortexDB is a NoSQL data base technology providing an unlimited platform for individual enterprise web applications at the price of standard software.

Data Sheet (2015) .pdf: Presentation_Cortex_150217

Download fact sheet (.pdf): CortexDB

Big Data: Three questions to CortexDB

Jan Buss, CEO, CortexDB, May 2014.

Q1. What is your current product offering?

Cortex AG powers CortexDB, a temporal, multimodal NoSQL database technology. CortexDB provides an unlimited platform for individual enterprise web applications at the price of standard software.

  • CortexDB is a temporal database technology. A temporal database includes valid time and transaction time. CortexDB also combines these attributes to form bi-temporal data. Temporal databases are more powerful than traditional databases that ensure only the truth at the transaction time and ignore the validity period.
    • Example: In master data management, users can add a validity date (“valid from…”) to the information in each field in a data set. That enables them to view and track back each change in any time-related context they choose. As a result, the database knows exactly how all the information has evolved over time, including both past and future values.
  • CortexDB is a multimodal NoSQL technology. It combines ALL the advantages of the various types of NoSQL databases (Key value; Document store; Graph DB; Multi value DB; Column DB). This is achieved via “CorAIT” (Active Index Transformation). That means that there is no special or user-defined index used for fast data access, the content is stored in schema-less mode inside CortexDB. All queries to the database run at the same speed as index databases – for all fields, combinations of fields, and even for linked information.
    • Example: When exploding a bill of materials for generic structures – such as in the automotive industry – this technology makes it possible to search recursively structured data quickly and efficiently, using any reference chains or attributes required. This enables users to determine the bill of materials for the vehicle concerned. A real-life case like this usually takes several seconds – or even minutes – but with CortexDB the result is displayed in a few milliseconds.
  • CortexDB is a distributed database technology that runs on Linux, Windows and MacOS, The CortexDB has also been ported to Android and ARM systems like Raspberry Pi.
    • Distributed databases use master/slave synchronicity. A slave server only receives filtered data, enabling dedicated servers to be brought in for special tasks. This ensures data integrity and increases security.
  • CortexDB includes a sophisticated security concept that can be activated automatically, even for in-house applications. Software developers can take advantage of this function, saving them the extra work in their own source code.

CortexDB provides unique technical features delivering various benefits:

  • Flexibility to change the database schema as required by business departments and software developers – the system adapts to the processes rather than the other way round
  • Extremely high database performance on standard hardware (low footprint)
  • Rapid and agile application development across innovative data services without programming
  • Simple modeling of complex structures
  • Change requests on the fly, enabling self-service usage by business units

Cortex AG sells exclusively through partners. Cortex supplies the technology, our partners provide sales and services. Marketing is shared between Cortex and its partners.

Cortex partners act either as value added resellers (VARs) or technology partners who are interested in using CortexDB inside their own products.

Cortex AG delivers server capacity based in the datacenter of a German provider, enabling customers, partners and Cortex itself to run CortexDB (and applications based on it) as cloud solutions. Large enterprises and other companies wishing to operate the database in-house are free to do so.

CortexDB is often used for hitherto unsolved database problems and for those that are very complex to solve. However, its ease of use has led to companies increasingly adopting it for “simple” tasks and new developments.

Developers who are frustrated by the limits imposed by conventional relational databases (capacity, complexity, speed, costs, amount of work involved etc.) find that CortexDB makes their everyday work significantly easier once they have learned how to use it.

As multiple tasks can be solved by the database itself and by simple configuration settings, developers only need to concentrate on their real work, not on optimizing database queries, the permissions system or other system-related tasks.

Departments with problems that can usually only be handled by the IT department find that CortexDB and applications based on it are the right self-service tools to help them work autonomously. One of the main reasons for this is that configuration settings can deal with multiple tasks and queries that previously required the assistance of an administrator.

CortexDB is used by midsized and large companies across all industries for a wide range of different purposes. They include both small and large amounts of data with complex structures.

Some solutions based on CortexDB:

  • Coordination of bills of materials between construction, logistics and production
  • Disposition of staff, equipment and orders in the surveillance industry
  • Validation of multilingual master data at a pan-European seed production company, in order to establish the comparability of similar products between different countries
  • Interactive utilization for data capture and semi-automatic data administration in the insurance industry, in order to improve data quality and maintain it long term

Q2. Who are your current customers and how do they typically use your product?

The go-to-market stage started at the end of 2012, and Cortex AG already has several automotive customers, including BMW, Mercedes-Benz and Volkswagen.

Cortex also has experience in the manufacturing and energy industries as well as in the service sector, such as insurance companies. Small and midsize companies can also benefit from CortexDB, as our pricing meets the needs of SMEs.
CortexDB has typically been deployed for predictive analytics and BI, performance management, master data management (MDM), unifying bills of materials (BOM) with complex, recursive data structures, disposition planning, and product lifecycle management.

One example is test vehicles. These comprise approximately 15,000 parts with between 1030–1060 configuration possibilities. This high variance is due to the number of vehicle models, engine types, displacement sizes, optional equipment, interior fittings and colors. To obtain a single car from these options, the bill of materials needs to be “exploded” recursively. Many other parameters can have an effect here too, such as validities and assembly stipulations, which increase the level of complexity even further. Interpreting these structures with CortexDB is fast and efficient, and users receive the results in a matter of milliseconds.

Q3. What are the main new technical features you are currently working on and why?

We already implemented Google’s V8 JavaScript engine inside CortexDB to perform calculations during the database query and calculation of pivot tables, which form the basis of the dashboard functionality. This approach will be extended so that program code and data are held together in the database.

Server-side implementation of Google V8.

The server is already capable of executing JavaScript on the server side today. This is used for a number of things, including on-the-fly results within reports that are not stored in the database at the moment they are generated. This enables users to add any calculation formulae they wish within a dataset. When they prepare their report, the linked structures are then fully resolved and the calculations executed. Users can of course save the results of the calculations in the database if they so desire.

Users can then perform further calculations with the results. For example, they could use real values to obtain extrapolated results and save them. If these are then used as historical information – in other words, if a validity date has been saved in the field together with the data – the user can compare the actual situation with the calculated values. This could be used to compare forecast figures with actuals.

In the future, it will also be possible to generate (partial) scripts via JavaScript and save them in data sets, which in turn generate results or other script content themselves. The goal is to produce self-optimizing programs that can make a forecast more efficiently and effectively based on the available data. This will also improve semantic analysis and facilitate the analysis of data that are still seen as unstructured or unmanageable.


Sponsored by Cortex AG

You may also like...