THE ATOM PLATFORM

We have prepared our own model for the development of mass capacity systems based on the observations gathered during a large number of implementations. The core of such systems is our in-house software framework, which can be used to develop very efficient and scalable systems for the most demanding clients.

The solutions are based on a unified software framework (i.e. a platform that is identical in multiple installations) on which micro-services are run (once called by us, before the concept of microservices was create – modules or converters), developed specifically for our clients (i.e. modules that differ depending on the installation). This means that, on the one hand, we do not develop our systems from scratch, because we are using a stable and tested platform that has been proven in many implementations. On the other hand, this approach allows us to best adapt to our clients’ needs and create solutions optimised specifically to match their expectations.

Our in-house Atom platform is our pride and glory. It was created specifically for high performance and scalability, in order to build sophisticated systems, particularly mass capacity systems.

Suffice it to say that the Atom platform will typically be capable of processing tens of thousands of transactions per second in the grid, even when using perfectly ordinary hardware. With accordingly designed data structures and hardware designed for operation in a grid, the Atom platform can process as many as hundreds of thousands or millions of transactions per second.

FIND OUT MORE ABOUT THE ATOM PLATFORM:

The idea of the atom platform MORE

The Atom platform (which is also referred to as the transaction and payment platform or PTR, due to the area it is frequently implemented in) is our own proprietary solution, developed by a team of analysts, engineers and programmers – graduates from technical faculties of the best universities.

Atom is the result of over a decade of research and work on mass capacity systems. The first version was developed already in the 1990s. Since then, the system has been repeatedly implemented at many companies during the following years, from financial institutions (banks) through capital market institutions (stock exchange, brokerage firms) and service or retail businesses to production companies.

As a mass capacity system, the Atom platform is a finished and stable computer system capable of running hundreds or even thousands of micro-services and combining them into more complicated tasks or groups of tasks.

Atom is the classic mass capacity system. This means that it has been designed specifically for lightning-fast processing of very large data sets. The whole design of the system, programming techniques, algorithms and operating solutions have all been selected specifically for high performance.

The Atom platform controls its servers and creates something resembling a cloud or grid of servers, enabling the configuration and running of processes on all machines as if they were a single virtual server. This means that there is a central point capable of configuring the entire system, managing its functioning and monitoring its operation.

Naturally, Atom can also function in a fully virtualised or containerised environment – depending on the client’s expectations

The Atom platform does not require expensive hardware. It can successfully run on ordinary, cheap and commercially available servers, forming efficient computing clouds.

The Atom platform can be used to create any system thanks to its performance, scalability and integrated management mechanisms.

Origins of the atom platform MORE

There are many different methodologies that can be used to implement IT systems. Unfortunately, the great majority of these methodologies are quite useless when applied to the mass capacity systems we develop. Their shortcomings can be particularly noticeable in certain implementations. The rule is simple – there is no point in using ordinary solutions if you need maximum performance and scalability.

During the many years that have passed since the founding of our company, we also considered various models for the development of mass capacity systems:

Tested development model	Our observations
Creating a single “off-the-shelf” application suitable for all clients, without any restrictions on its functionality	We found that it was impossible to prepare such a readymade application. It would be too complex for some of our clients and too trivial for others. Due to the different requirements and expectations, it was completely impossible to find a common denominator suitable for all clients. Moreover, such an application would certainly not be a mass capacity system – it would not be able to process all kinds of data in an optimised way. The configuration of such an application could also be rather time-consuming. Let’s not forget that we wanted to avoid the common situation where the configuration of some readymade systems takes just as much time as the development of a system “from scratch”.
Creating an “off-the-shelf” application with restricted functionality to only meet the basic requirements of the clients	This solution would also be rather pointless because the optimisation of a mass capacity system requires the construction of data structures. They may be different in each case. How, then, can you develop even a trivial off-the-shelf system without data structures? You can’t.
Developing custom systems	This is our preferred approach because that is precisely the way we have implemented our systems since 1995. This method can be used to develop very efficient solutions adapted to specific needs. Consequently, this approach can meet the expectations for mass capacity systems. Plus, the clients pay only for the features they actually use. Unfortunately, every such implementation is unique. Errors found in one system will be completely different from the errors found in a second system. This means that all patches (including upgrades) will have to be prepared separately, increasing the costs of broadly defined application support and maintenance.

Comparison of system development methods

The table below shows a comparison of different application development methods.

	Off-the-shelf application with basic functionality	Extended off-the-shelf application with full functionality	Application developed and implemented upon request	Custom application using a software framework
Performance	“-”	“-” extensive configurable environments require a lot of extra resources to enable such configurability – at the expense of performance	“+” excellent – the system is designed to meet all performance requirements of the client	“+” the software framework ensures maximum scalability and performance (grid/cloud). Both the data structures and the modules themselves can be optimised as necessary
Functionality	“+/-” very good for basic duties but insufficient for more complex implementations	“+” sometimes even too much	“+” precisely matching the client’s needs	“+” precisely matching the client’s needs
Ease of implementation	“+”	“+” in extended, configurable systems, the configuration process is a separate body of knowledge in its own right, and it may (but does not have to) be as difficult as developing the entire system from scratch; the client may also be forced to adapt some of its solutions to match the system	“-” unfortunately, it is necessary to implement the entire application, including all standard mechanisms as well, but, on the other hand, the application will be customised to the intended company – it will be more suitable and easier to implement (e.g. the training will not concern a new methodology of approach to the system – it will only cover the use of the system)	“+” the platforms we use are readymade. Including a standard user interface. All that is left to do is to create the body of the system
Time of implementation	“+” outstanding	“-” in some cases, configuration of a large, configurable application may be so complex that it takes as much time as developing a suitable system “from scratch”	“-”	“+” much shorter than if the system had to be created from scratch – the platforms are available immediately. It is also easy to use existing configurations
Clients pay for what they actually use	“-”	“-” the client pays for the application anyway, and if a certain option is undesirable, the client pays for it in the standard price and then pays for its removal during configuration	“+” this is one of its primary advantages	“+” clients always purchase the entire platform, but they almost always use all of it
Availability of updates / code sharing with other users	“+”	“-/+” (partial – depending on how much the configuration changes the core of the system)	“-”	“+” regarding the platform – full availability; regarding the body of the system – yes, if the particular body of the system has been installed for multiple clients

Design of the atom platform MORE

Design of the atom platform

The solutions are based on a unified software framework (i.e. a platform that is identical in multiple installations) complemented by micro-services developed specifically for our clients (i.e. modules that differ depending on the installation). This means that, on the one hand, we do not develop our systems from scratch, because we are using a stable and tested platform that has been proven in many implementations. On the other hand, this approach allows us to best adapt to our clients’ needs and create solutions optimised specifically to match their expectations.

Supported operating and database systems

Modules for the Atom platform

Grid controlled by the Atom platform

Supported operating and database systems

The entire system was written in C++, and it was separately optimised for operation in Windows and in Linux.

Furthermore, the system was designed to enable simultaneous maintenance of the source codes for both environments because the great majority of the source code is transferable, which makes it easier to provide support for these key operating systems.

The system can also run in heterogeneous environments, controlled simultaneously by the two above-mentioned operating systems. In such a situation, however, the modules run on the Atom platform should be created separately for each operating system to ensure maximum performance.

The system uses two categories of database servers. One server type is used for the internal configuration of the platform itself, and the second is used for the physical processing of data. These servers can in practice include any major brands on the market – from MS SQL Server to solutions such as mySQL or PostgreSQL. For maximum performance, we recommend using the dedicated VoltDB database server, but the final choice is, obviously, up to the client.

Through cooperation of VoltDB with the Atom platform, we can provide scalability unlike anything that can be achieved using other methods. We can present solutions capable of processing millions of records or even entire transactions per second. Highly complex operations can be handled as well.

VoltDB with the Atom platform also enables unparalleled scalability. In the case of a problem with performance, mass capacity systems using Atom and VoltDB can be significantly upscaled.

Modules for the Atom platform

The key term relevant to the Atom platform are modules (implemented in the form of micro-services). These are fairly small and simple compiled program blocks created at our company, usually in C++, in order to ensure maximum performance. In practice, the modules can be created using any programming language.

Two types of modules can be run on the Atom platform:

on-line modules, which process data all the time, without any delay,
off-line modules, which are run at specific intervals and process data for which immediate processing is not necessary (i.e. for which a one-minute delay is acceptable).

The modules perform simple, atomic operations. This makes them relatively easy to program and, more importantly, test.

Let’s assume it is necessary to develop a system that is going to receive data from an indicated IMAP e-mail account. Then, the attachment to the e-mail has to be decrypted using the AES algorithm and decompressed, and the resultant CSV file has to be imported into the database. Next, after aggregate calculation, an XML file has to be created, compressed, encrypted with the 3DES algorithm and sent to the indicated address of an FTP server. In the described case, it will be necessary to build the following modules:

Module to download data from the indicated IMAP server;
Decryption module (AES);
Decompression module (zip);
Module for import into the database;
Aggregate calculation module;
Module for export into an .xml file;
Compression module;
Encryption module (AES);
Data transmission module (FTP).

As evidenced by this example, modules 1, 2, 3, 7, 8 and 9 do not have to interfere in any way with the logic of the data they transmit. Consequently, they may be run repeatedly in different contexts for completely different data.

Changing the data aggregation method only requires a modification of module 5. All other modules may be left unchanged.

If the task described above requires a change of the method used for communication with the target system, e.g. transmission of data using a dedicated protocol, it will be enough to simply replace modules 6–9. All other components may be left unchanged.

The modules are a kind of a plug-in. That is also how they can be installed on the Atom platform, which distributes the executable code of the plug-ins to all of the computers it controls.

Processes

The module run on the Atom platform receives a specific configuration. A module with such a configuration is referred to as a process. This means that, for instance, a data decryption module can create various processes – it may be started with a completely different configuration each time (different passwords, different file locations, etc.). Consequently, a single module can be used as a component of an unlimited number of processes.

This obviously makes it easier to the test the system created using the Atom platform, and it enables actual reusability of the same source code in different business applications.

Tasks

By definition, the processes are too simple to directly fulfil the real requirements of today’s businesses. However, the Atom platform can establish connections between the processes to create entire graphs of associations.

In the example shown in the “Modules on the Atom platform” section, the graph is fairly uncomplicated – it is simply an ordinary sequence of processes from 1 to 9. However, there could be a special process initialised only if the decryption of the file fails or a process used to transmit the data with a different method if the FTP server is unavailable.

The graph of associations between specific processes is referred to as a task. The task can actually be as complex as may be needed, and it may contain any number of processes. The processes can be associated in any way (e.g. they may be run sequentially, cyclically, etc.). Every task only requires the identification of the first process to be run. The Atom platform will make sure that the processes are run in a suitable order, and, naturally, it will fully monitor their results, faults, etc.

What is more, the mechanism works both with off-line tasks (producing a conventional but very efficient batch job mechanism) and on-line tasks. The latter is particularly interesting because it means that the Atom platform can be used to define a sequence of on-line connections between processes running simultaneously. In such a situation, the graph of associations will no longer be a simple illustration of the sequence of the associations – it will indicate which processes send queries to which processes. The Atom platform can then control the loading of the processes by running them accordingly on new machines (or by closing useless processes).

Task groups

Configured tasks are combined on the Atom platform into larger structures referred to as task groups. This makes system administration much easier. Separate Beginning-of-Day (BOD) and End-of-Day (EOD) tasks can also be defined for every task group.

Time zones

The system obviously supports different time zones, all at the same time. Specific task groups can, therefore, be assigned to a specific time zone, so that the time and method of processing can be defined depending on local conditions. This means that the Atom platform will be a perfect solution for all global companies operating in multiple time zones that require central processing of specific data.

Grid controlled by the Atom platform

The Atom platform is a complete solution capable of creating something resembling a cloud or grid running on a server farm. The Atom platform controls these machines (physical or virtual) and manages their structure as if it was a single system.

The platform enables fully central management of the system, which can not only be configured (including central installation of modules) but also monitored.

The Atom platform, as a complete tool dedicated to data processing, includes the following:

configuration system (AtomConfig) – used for central creation of any business process based on task groups, tasks and processes developed using dedicated modules,
monitoring system (AtomConsole). All of the key aspects of the operation of the platform are monitored, i.e. the system reports both the parameters of the physical machines and the basic aspects of data processing. The data are available via a dedicated administration panel,
load management system (AtomLoadBalancing). This is one of the key components of the platform. The system actively manages processor workload to optimise the operation of the entire platform, enabling maximum use of its hardware capacity and prioritising of specific processes through the assignment of extra resources.

Advantages of the platform MORE

Advantages of the platform

Our system was developed in close cooperation with banking and mass data processing on capital markets, which is why it was designed to accommodate at least three types of system growth:
– “normal” growth, involving the organic increase of load corresponding to the normal development of the economy,
– rapid growth, where the client decides to manage a brand new process (e.g. incorporation of a large group of new accounts to be processed) or acquires a large group of customers as a result of a merger and includes such customers in the service,
– the third type of growth may be caused by an increase of the complexity of algorithms, which, however, does not necessarily have to increase their processing time.
Atom is a continuously developing system. Our clients are now offered access to version 4.0 of our platform. Moreover, all upgrades are completely free of charge. We also offer the full training cycle covering system management, configuration of tasks, processes and task groups, training on system monitoring, etc. This way, our client or local partner can successfully manage the system developed on the Atom platform on their own, without our assistance.

Price – including both software and hardware:
The system was designed to run on regular hardware. This means that, instead of using a single, expensive server, we propose using multiple servers, which can even be much cheaper overall. This approach can:
- reduce the overall costs of system operation and maintenance (total cost of ownership),
- mitigate a number of risks, including the near elimination of one of the main risks, i.e. the “single point of failure”.
The Atom platform is a special-purpose solution. This means that it is not designed to simply enable our clients to catch up with their competition but is intended to provide a genuine competitive edge.
The Atom platform is designed for use in any area where large amounts of data have to be processed. This makes it a universal platform for data exchange between heterogeneous systems, which can be used in all of the following areas:
- in banking (transaction banking, trade support banking, card transactions),
- on capital markets (as the integration layer for the stock exchange transaction systems),
- as support for IT services, e.g. as support for the integration of electronic customer contact channels,
- as a component of near-real-time systems.
The Atom platform has been designed to:
- support streaming data (capital markets, IT),
- support static data, e.g. files, stored procedures, etc. (banking, IT),
- support all known communication protocols,
- support all known encryption and data protection protocols,
- support all known data compression mechanisms,
- support any type of data in any format.
Atom is a mass capacity system, which is why speed is its primary priority. This is achieved through the optimisation of the platform on 4 primary levels:
- code optimisation – processing modules have been continuously optimised for many years in order to achieve the maximum possible processing speed,
- communication protocol optimisation – thanks to many years of studies on data transmission, we now use the most efficient and fastest data transmission mechanism,- processing optimisation – using 100% of the capacity of the physical machine,
- using Atom database solutions – including, in particular, the partitioning of the entire database (instead of individual tables).
Atom is a complete solution dedicated to data processing. It has a set of tools used for system configuration, installation of routine system monitoring and identification of problems (extensive log system).
In our past systems, we have used database solutions of all major providers, i.e. Oracle/Microsoft/PostgreSQL. Upon the client’s request, we can also optimise the system for other databases, e.g. in-memory databases (Oracle TimesTen, MonetDB).
However, due to reasons of performance and scalability, we strongly recommend using the VoltDB database with the Atom platform for any important implementations.

The Atom platform can also be easily implemented on various operating systems, including systems of the Microsoft product group as well as Unix/Linux.
We usually describe the Atom platform with the metaphor of the periodic table, i.e. a fundamental set of elements (modules performing simple, basic actions, such as compression, encryption, aggregation, etc.) that the clients themselves use to build more complex “chemical compounds”.
Atomic modules can be easily implemented and tested because they are clearly separated from other modules. Atomic modules are also easier to design and implement. They are also simply more cost-effective at all stages of development. It is also easier to modify the modules because the programmer only has to solve a very local problem concerning a small amount of source code, which is much faster than an analysis of extensive procedures.
The atomic design can also significantly increase the stability of the entire solution because the tasks are based on compiled and already tested atoms (modules). The person configuring the process or process task cannot interfere with the compiled code.
The tested code is repeatedly reused in a compiled form. The “Reusability of the code” requirement is therefore not implemented based on classes or libraries in the particular programming language but based on a compiled code that can interact with other similar modules using the defined interfaces. Consequently, reusability significantly reduces the cost of implementation of further services on the particular instance of the Atom platform.
Stringent standardisation of code is one of the principles of atomic design in order to create large programmer teams and make it easier to expand such teams and share knowledge within each team easily and efficiently.

“Unattended” operation

Unattended operation means that there is no interaction with the user during the running of the tasks. This is simply the embodiment of the concept: “set once, perform many times”. This way, there is no need to manually interfere with the system during its operation. In case of a fault, the system will initiate suitable actions itself, based on pre-defined scenarios. This approach minimises the most frequent cause of major faults – hasty and careless human behaviour in response to a non-critical fault.

Support by a large number of programmers at our company

The concept of atomic design and the resultant imperative of code standardisation significantly increase the safety of the entire solution because there are many programmers capable of modifying the simple, atomic modules. Most interventions do not require any special knowledge of the product and its mechanisms, e.g., factoring – all that is needed are the skills required to make simple programming modifications. This enables parallel development and, consequently, thanks to our large team of programmers, significant acceleration of system implementation in comparison with solutions of the competition.

Development of the atom platform MORE

Atom is the result of over a decade of our research and work on mass capacity systems. The first version was developed in the 1990s in order to answer the specific needs of our clients. Since then, the system has been implemented at many companies, from financial institutions (banks) through capital market institutions (stock exchange, brokerage firms) and service or retail business to production companies.

The Atom platform will be well-suited for all areas where it is necessary to handle complex processes and process enormous amounts of data in the shortest time possible. We have managed to achieve this by listening to the suggestions of our clients and thanks to our own continuous internal development work.

The Enterprise Edition of the Atom platform has been created for the most demanding companies, for whom system performance, stability and high availability, as well as continuity of business (High Availability / Disaster Recovery) are necessary to fulfil their business objectives.

In the Enterprise Edition, the Atom platform is intended to handle business and operating processes involving the processing of very large amounts of data from multiple distributed sources (Big Data) as well as processes in which the data are processed in order to make split-second decisions, in the on-line mode (Fast Data). The architecture of the Enterprise Edition is entirely dedicated to extreme performance and processing speed. This is achieved thanks to the integrated database engine, which, where necessary, can process partitioned data at a rate that cannot be achieved by standard database systems.

Since the Enterprise Edition is usually dedicated to critical processes, it has been enhanced with additional mechanisms to ensure high availability – this was achieved, in particular, through the redundancy of all of its components.This version also ensures a level of reliability unequalled by the standard solutions on the market, guaranteeing business continuity and optimisation of the Platform for automatic restart in case of a major fault (Disaster Recovery) – these mechanisms are intended to guarantee 100% transactability, even in case of a failure.

The above-mentioned features enable the achievement of unprecedented levels of data coherence and security, which are required to process sensitive data and, consequently, to make real-time decisions (which result in the immediate creation of business obligations) based on such data.

The table below shows the adopted roadmap for the development of the Atom Enterprise Edition:

Parameters	Atom v.3.n (current versions)	Atom v.4 (soon)	Atom v.5 (future)
Log	SQL database (up to 100/s)	super fast queue (short term) SQL database (long term)(up to 30 000/s)	super fast queue (short term) SQL database (long term) (up to 30 000/s)
BigData (Volume)	+	++	++
BigData (Velocity)	+	+++ (on-line/real-time)	++++ (on-line/real-time) with transactions
BigData (Value)	+	+++ (on-line/real-time)	++++ (on-line/real-time) with transactions
BigData (Variety)	hard coded	hard coded	data independent – binary XML
On-line processing	limited	unlimited – massive parallelism	unlimited – massive parallelism plus distributed transactions
Adaptive systems	–	+	++ (speed)
Reporting modules	–	++	++
AI	–	+	++ (speed)
On-line pull converters	+	++	++
On-line push converters	–	++	+++ (with queues)
Queue converters	–	–	+ (even more: all others simulated by queue converters
Data queues between converters	–	–	+ multiple in & out queues

RELIABILITY OF THE ATOM PLATFORM MORE

RELIABILITY OF THE ATOM PLATFORM

The high quality of the platform is assured thanks to our long experience with the design and implementation of solutions supporting mass capacity systems, particularly systems for the financial markets, and by the industry awards, such as Europroduct 2012 and recommendation of Gazeta Bankowa in the technology competitions Hit of 2011 and Hit of 2010 for Financial Institutions in the “Solution” category.

1) Possible use casesRozwiń

Possible use cases

System integration

Mass processing of on-line transactions

Modelling of business processes

Reporting

System integration

The Atom Platform is frequently used as an EDB (Enterprise Data Bus) solution, functioning as a universal translator between different systems. That is why one of the critical issues was for it to be easily interfaced with any number of heterogeneous data sources via any communication channels.

The Atom platform has been designed to:

support streaming data (capital markets, IT),
support static data, e.g. files, stored procedures, etc. (banking, IT),
support all known communication protocols,
support all known encryption and data protection protocols,
support all known data compression mechanisms,
support any type of data in any format.

From the very beginning, the Atom Platform was meant to ensure that interfacing with a large number of systems that send or receive data to or from the platform via all known channels in all known formats is as easy and simple as possible. The data may be compressed and encrypted.

Sources of data may include any systems used to store data (warehouses) or OLTP data processing systems, which means that Atom will be suitable both for financial and capital markets and in the sector of retail chains and international business organisations, where the management of data (e.g. financial data) requires the integration of data sets from multiple countries, generated in different time zones and based on different calendars. That is why Atom is the tool for global integration of data that shines the most in heterogeneous business organisations.

In the case of data from multiple heterogeneous sources, the key aspect is to prevent the risks related to the loss and/or misinterpretation of data. That is why the platform can be provided with a number of modules used for validation, clearing and deduplication of data as well as tools for defragmentation and consolidation, conversion, reconciliation, matching and aggregations of any kind.

Modelling of business processes

The Atom platform has very convenient tools for the modelling of business processes.

Every process is constructed from simple, configurable modules used to perform an atomic action (data decompression, data validation, aggregation, data encryption, etc.).

We describe this metaphorically by stating that the engine of the system is delivered complete with the entire “periodic table”, i.e. a series of modules performing simple business functions (atoms), which can be combined by the client to create very complex “chemical compounds” reflecting the business processes. The system has a business process modeller, and, at the same time, the entire solution is faster and more efficient than known systems by other providers because we do not use script languages.

Since time is critical in mass data processing, our solution has also been provided with a range of tools for temporal modelling of the processes. We can specify the time of commencement and completion as well as the duration of the individual data processing processes.

Mass processing of on-line transactions

The typical use of the Atom platform is the mass processing of on-line transactions, which appear in large quantities and must be processed in a short time. This refers to, for instance, payment card authorisation transactions – sale transactions recorded directly at the POS (point-of-sale) and requiring the approval or decline of authorisation – or, in completely different applications, information received in real time from various sensors that require a response of robots on the production line.

In such transactions, the critical issue is processing time. Naturally, this entails more than simply recording the transactions in the database – it is necessary to perform the entire process, including suitable calculations, aggregation, communication with other systems, etc. Obviously, a customer at the POS has to be served in the shortest possible time, and robots on a production line cannot wait for the decision of the computer system on how to proceed.

Modelling of business processes

The Atom platform has very convenient tools for the modelling of business processes.

Every process is constructed from simple, configurable modules used to perform an atomic action (data decompression, data validation, aggregation, data encryption, etc.).

Time is critical in mass data processing, which is why our solution has also been provided with a range of tools for temporal modelling of the processes. We can specify the time of commencement and completion as well as the duration of the individual data processing processes.

Reporting

A system built on the Atom platform is an excellent way to process data used in reporting systems. Obviously, there is no point in using Atom to visualise the data (although this, too, could be done with a suitable module). The platform can, however, successfully process enormous data sets. This way, you will be able to obtain summaries almost immediately after a specific economic event occurs instead of having to spend a large amount of time feeding the data into the system (as is the case with regular data warehouses).

Time is a critical factor in today’s world, and speed is and will continue to be the primary contributor to competitive advantage. That is why we believe that the key is not the type of data received as a result of our analyses (because the competition will be able to generate the exact same data) but, most importantly, how quickly you can generate them.

Reporting on the Atom platform can be performed entirely on-line. Modules run on the Atom platform can routinely perform complex calculations and control other systems: they can calculate the customer’s creditworthiness, evaluate the behaviour of customers browsing a website or analyse the customers’ shopping choices made at just that moment.

2) Sample applicationsRozwiń

Sample applications

The Atom platform is dedicated to data processing, which is why it will be perfect for all areas where the business needs concern the processing of operating and analytical data.

Wherever there is a business need for mass data processing, integration of multiple sources and formats of data, various aggregations, generation of summaries, reports and other data input for management, MIS and ERP systems, etc., the Atom platform will be the optimal solution. So far, the Atom platform has been used in the following products and services for the banking sector and capital markets:

Service addressed to the corporate customer, designed for the automatic accounting of receivables. It can be used for efficient management of the collection and recording of receivables. The system generates the specified reports in the native formats of the client’s financial and accounting systems.
The system is designed for bank clients that wish to improve and automate the routine coordination of their receivables by minimising the number of questionable payments, i.e. payments with incorrect or incomplete data, which cannot be automatically approved by the bank’s client.
The system uses pre-registered files to generate an electronic “Debit request” sent to the payer’s bank via the system. Based on the registered response, it generates feedback to the bank’s client in the form of an electronic report imported by the creditor into its system. From this moment, further invoices of the given payer can be paid via a Payment Order.
The cash pooling system enables banks to provide their clients with a financial service where the account balances of companies belonging to the same holding company are treated collectively. This can be used to concentrate the funds of the companies in a single account of the group and manage this amount. Temporary financial surplus of some of the entities compensates for the shortages of others. To the holding company, this reduces the costs of interest paid by its subsidiaries with negative account balances by providing funding from the subsidiaries with positive balances. The system supports real and virtual cash pooling.
The system can perform two basic functions because it is based on a range of indicators that describe transactions, products, and balances of the customer. As a contract monitor, it can be used as an “early warning” system to deliver data to the consultant showing if the customer is complying with the contractual provisions.

The same set of indicators, however, can be used to create a transparent scale of rewards and incentives for the customer in a loyalty programme.
The system can be used to perform evaluations at specific intervals and notify customer consultants and the management about the results of the rankings.
The system is a very flexible tool designed to create a fees policy and quickly respond to the changing conditions on the market. The system can be configured to create tables for customers, products, transactions, and cash operations. It can also report the effectiveness of the implemented policies in reference to units of time and with breakdown into individual products.
This system prepares a set of reports and delivers them to the client in the format selected by the client or in the default format. The system can deliver treasury reports (Daily Cash Reports) and supply the financial and accounting systems of the client with data. Many other reports can also be prepared for the client’s financial departments because the capabilities of data collection, aggregation and validation are unlimited.
The Atom platform is an excellent tool for creating reporting systems because it has mechanisms that can import data from virtually any source, running at different locations and on any operating systems or machines. The data may be available both locally and remotely (via the Internet or dedicated VPN connections). Therefore, Atom will be well-suited for use as an integration platform to combine data from heterogeneous sources.

The Atom platform proposes a hybrid approach – using both OLAP and OLTP databases. As a result, the platform can not only generate typical cross-sectional reports (as in a conventional warehouse), but it can also report specific, individual transactions (as in OLTP systems). A single report can include both of these types of information. Such reports can be prepared with unrivalled speed because the data on the Atom platform have already been validated, transformed and supplemented. Data quality will also not be an issue in this case. This way, the Atom Platform can be used to create very complicated and very quick reports. The reports are prepared automatically at specific times. They can be made available in the form of services (e.g. WebServices) so that other IT systems can import the data reported by the platform in the on-line mode. The Atom platform can generate suitable reports and distribute them, e.g. as a PDF file attached to an email. The reports are prepared without interaction with the user – this method is particularly effective when used with reports that require complex processing.
A system based on the Atom platform is used as the integration layer in the most important institution in the Polish capital market. Its primary function is to download the stream of messages from the transaction system, make the necessary aggregations and modifications and then send them to a number of servers “responsible for the picture of the market”. The purpose of the platform is to transfer the processing logic from the receiving systems to the layer in order to prevent complications related to the need for repeated adjustment to changes of the transaction system and changes within the system.

As the universal intra-system translator, the Atom platform also simplifies the architecture of the client’s systems. It is one of the most advanced and fastest data buses on the market. During peaks, its performance significantly exceeds the planned 5000 messages per second per low-performance server. Additional benefits include the capability to function both in the synchronous mode (receiving and sending data streams) and asynchronous mode (receiving and sending finite data sets) and also via a very diverse range of channels (web service, stored procedures, TCP/IP, other). The system is a comprehensive solution with interfaces for process configuration (AtomConfig) and real-time process monitoring (AtomConsole) as well as an extensive load balancing, monitoring and COB system.
The system will be well-suited for situations where electronic customer contact channels have to be harmonised and streamlined. It is designed to eliminate the problems that may occur when the number of contact channels (call centre, email or SMS) or the number of senders (transaction systems, CRM and marketing systems, debt collection systems, etc.) increase. This is normally addressed by simultaneously increasing the number of systems and extending their functions in order to enhance their communication capabilities, which may lead to problems caused by the duplication of communication solutions in successive systems, multiplication of billing systems, large costs of interface modification and absence of coordination in the customer contact area.
System designed to monitor balances and transactions and manage cash flows on multiple accounts of the corporate customer and forecast the customer’s cash level in the future.
System for secure two-way communication between a corporate customer of the bank and the bank itself. It is a data exchange platform designed for automation of communication between financial and accounting systems, ERP systems and HR systems of the bank customer and the bank system. The system enables the bank customer to send a package of transfers to be processed, check their processing status or download bank statements directly from an internal system, e.g. from a financial and accounting system, without the need to log into the systems of the bank.
System dedicated to the comprehensive management of instant transfers in cooperation with one or multiple providers of the instant payment service (e.g. Express ELIXIR, Blue Cash, PayU Express) and operators of the PayByLink service (including eCard, PayLane, PayPro and CashBill).
System for the management and processing of bank transactions (mass incoming and outgoing domestic payments and payment orders) capable of:
1. identifying and managing incoming payments,
2. managing outgoing payments / withdrawals,
3. managing cash handling.