EUREKA PROJECT > 4966 PULSE
PORTUGAL
Pulse: Real-Time Business Intelligence
The project aims to develop a Business Intelligence product capable of processing thousands of events per second, continuously producing updated business information. By using PULSE, companies can continuously monitor their businesses, reacting to new information proactively and in real-time.
Status > ANNOUNCED - 28-Jun-2009 Technological Area Market Area Start Date > 31-May-2009 Duration > 18 Months Participating countries > PORTUGAL, NORWAY | Main contactFeedZai - Consultadoria e Inovacao Tecnologica PhD Pedro Bizarro > Managing Director Organisation type > SME |
Project Description
Modern organisations are not blossoming with data; they are drowning in data. Thousands of critical business data events are produced every second, from customers, suppliers, data centres, sales, web servers, IT (Information Technology) department, and more. Most organisations fail to discover and explore new trends and warnings on time simply because they cannot process incoming data efficiently. Existing technology was not designed to process thousands of small discrete events from multiple sources, correlating those events, checking for pattern occurrences, and reacting in real time. Nevertheless, such functionality is now critical for the modern enterprise. Activities or systems like stock trading, retailing and distribution, internet gaming, online betting, fraud detection, computer security, factory production lines, air traffic control, automatic toll charging, water supply systems, emergency room monitoring, real-time engine tuning, and, even amusement parks, share something fundamental: they all need to collect raw data from distributed sources and produce actionable information in real time. Today they all develop custom, cumbersome, complex, expensive systems to deliver real-time information. The aim of the PULSE project is to research and develop a disruptive product in the area of real-time business intelligence. PULSE aims to provide continuously updated business information, to the second, allowing key decision makers to monitor their enterprises, react to new information in real-time, and maximise their businesses’ potential. This product will further develop and incorporate several technologies and research results produced by the founders of the proponent company (FEEDZAI), while they were research professors at the university, and will be validated by real-world use-cases by a financial and security Norwegian IT company provider (KANTEGA). In particular, this project is characterised by four key objectives: O1. To create a real-time business intelligence software product that has a disruptive innovation potential when compared to the current state-of-the-art in the area, worldwide. O2. To guarantee that the project is aligned with the effective needs of the market, driving its requirements and validation by real industrial use cases and users. O3. To generate new knowledge in the area of real time complex event processing, making sure that this knowledge is made widely available. Such can be achieved by publishing results at leading conferences and forums as well as protecting IPR (Intellectual Property Right), as appropriate. O4. To improve Europe's competitiveness in the area of Information Technology and, particularly, in the area of Business Intelligence. Pulse is being designed to have four significant advantages over the current state-of-the-art: i) a breakthrough user interface to define business patterns to monitor and process in real-time, enabling fast deployment of projects by non-experts; ii) a richer development language that allows users to ask questions they cannot currently specify in other products; iii) better performance than competing products due to use of accelerating hardware; iv) very easy integration with data sources, especially web-based, and other computer systems. These characteristics will fulfil the following goals: 1) Develop a Business Intelligence disruptive product with global reach. The product is disruptive because it will be used to develop applications faster, cheaper, and with better performance than competing offers. 2) Assure that the product is aligned with real market needs and validated with real use cases. 3) Explore new markets and generate original research - FEEDZAI and KANTEGA enjoy extensive collaboration with a set of partners working on different markets and with a partner specialised in producing original scientific outputs. 4) Improve Europe's competitiveness in the global market.
Technological Development Envisaged
Modern IT (Information Technology) systems consist of numerous pieces of software and hardware, communicating and processing thousands of events per second. Finding out what is really happening and when it happens is very challenging: seemingly unrelated messages arrive to the system at very high rates and from different sources. For example, if the average mobile phone call length in one area of the network is much shorter than the overall average, it is likely that the calls are getting disconnected because an antenna is not working properly. This implies less airtime sales, more unhappy customers, and increased load in customer support lines. The odd thing is that although the cell phone company has the information it needs to quickly pin-point the fault, it will probably only detect the fault days or weeks later. The problem is not lack of information. The problem is the inability to digest it in real-time. Currently, organisations that seek to process large volumes of information in real-time have immense difficulty in doing so efficiently. Not only is the problem getting worse, the only two known approaches are defective. The traditional approach extracts information from operational databases, into decision support data warehouses, and then generates periodic reports. For achieving this, most companies rely on traditional databases (e.g., Oracle, MS SQL Server) for implementing both the operational and decision support databases, and sophisticated reporting tools for generating reports and dashboards (e.g., MicroStrategy, Business Objects). Unfortunately, these databases and reporting tools were not originally designed to process in real time thousands of small discrete events from multiple sources, checking for pattern occurrences, and acting in-time. Typically, latencies when using this approach range from many hours to weeks, which is clearly insufficient for a real-time business environment. The other emerging approach is the assumedly named complex event processing (CEP) technology. CEP aims at providing real-time actionable information extracted from those thousands of events per second. Having only appeared in 2005 as an independent branch of event processing, there are now a small number of companies providing solutions to try to address the problem. Examples include Oracle, IBM, TIBCO, Aleri, Progress, StreamBase and Coral8. The main problems with current CEP products are: a) they provide a very low level of abstraction for information processing, not focusing on the real needs of business decision makers; b) due to their low-level of abstractions they are very complex and cumbersome to use, leading to development projects that take months to complete by highly trained technicians; c) they are very expensive. As an example of these problems, it is worth mentioning that to try to implement something as simple and common as a double-bottom pattern detector, for financial analysis, currently requires hundreds of source-code lines written by specialists.
Main contactFeedZai - Consultadoria e Inovacao Tecnologica PhD Pedro Bizarro > Managing Director Organisation type > SME |
Contribution to project
The organisation will contribute to this project with its product "FeedZai Pulse". The product produces continuously updated business information that allows decision makers to monitor their enterprises, react to new information in real-time, and maximise their businesses potential. This product will incorporate several technologies and research ideas developed by the company founders with research professors at the UNIVERSITY OF COIMBRA. "FeedZai Pulse" is being designed with three significant technological advantages over the state-of-the-art: i) The interface used to define business patterns to monitor is very simple and easy to use (reducing the time to deploy to a fraction of total time - about 6 times faster in our experiments); ii) Better performance than competing products due to innovative technological breakthroughs; (initial comparisons with alternative offers have FeedZai Pulse show an improvement of 5 to 8 times in performance) and iii) Very easy integration with data sources and other computer systems. The product "FeedZai Pulse" implementation corresponds to FEEDZAI's intellectual property.
Expertise
The core business of FEEDZAI is the generation of business intelligence in near-real time by integrating information from multiple data sources. Building the FeedZai Pulse will require expert knowledge in databases and stream systems, event processing systems, distributed systems and integration, business intelligence, and in compilers and programming languages. FEEDZAI founders have expertise precisely in those areas. Pedro Bizarro, with a PhD specialised in databases and stream management systems (with publications in leading database conferences and journals - SIGMOD, VLDB, CIDR, TKDE), is currently the global co-leader of the Event Processing Technical Society (EPTS) Use Case Workgroup and leader of BiCEP - an event processing research project. Paulo Marques, with a PhD in distributed systems has done extensive research in programming languages (co-author of the most sold C# programming language book in PORTUGAL), created the first Enterprise Application Integration advanced course in PORTUGAL, and has worked in several Business Intelligence projects for the ESA (European Space Agency).
Main contactKantega Mr. Jon Oyvind Eriksen > General manager Organisation type > SME |
Contribution to project
The organisation will bring to this project its experience in the financial and security services sector, namely real-time banking fraud detection. KANTEGA’s responsibilities in the project are to specify and implement two use cases for the usage of "FeedZai Pulse", developed by FEEDZAI LDA., in the financial services market and perform the corresponding validation. These two use cases and the implementation correspond to KANTEGA's intellectual property.
Expertise
KANTEGA is a specialist in internet security solutions. Based on extensive experience with the banking industry, KANTEGA has developed market-leading security solutions. As part of the focus on security, KANTEGA has developed software for forensic investigation of cyber-crimes, including real-time event processing software to detect and prevent intrusion attempts and other attacks (including use of trojans) on the internet customer service infrastructure of a large financial services group (internet bank). KANTEGA builds customer self-service solutions and internet portals with support for Straight-Through Processing based on eID and digital signature. KANTEGA’s security solutions portfolio includes software for strong authentication, access management, federated identity management, single-signon and real-time monitoring and prevention.