Arkadiusz Roszczewski Email & Phone Number

Warsaw, Pl

New York, New York, Us

Key areas:Scoring modelsIFRS & US GAAP provisionsBaselKNF & OCC reportingCredit Risk MIS & BI

Warszawa, Mazowieckie, Pl

Direct supervision of a team of analysts and database developers.Architect and lead developer of reporting system of operations division. Responsible for various stages of reporting and BI projects: business analysis, system analysis, design and development of database and ETL process, statistical analysis and data mining, result visualization and OLAP including real-time solutions, insights and recommendations for business users.Directly involved in bank merger and system migration projects.

Brussels, Brussels Region, Be

Managing a team of analysts / developers, providing quantitative analytics and custom software for Personal Finance Operation Division.- Architect and lead developer of reporting infrastructure for Collections Department (ETL, datamart, analytics, presentation / visualization, control flow monitoring, end-user thick client design).- Introduction of new collections system: analytics / tests.- Designer and lead developer of customizable, real-time dashboard software.- RAD prototyping.- Ad-hoc analytics and modeling.- Technical cooperation with IT department and external software vendors.

Datamart and analytical tools design. Risk modeling and quantitative analytics. Member of decision engine working group - business analysis, RAD prototyping, audit of system architecture provided by software vendor.

The first six months of the employment period consisted of hands-on trainings performed in London in the area of credit risk analytics of derivatives portfolio. Current key responsibilities on this position are:- Design of architecture of business intelligence platform for Product Control department including the architecture of analytical tools and data structures for credit risk monitoring framework.- Calculation of credit risk reserve level of derivatives portfolio including analytics and monitoring of credit risk parameters related with exposure levels, portfolio segmentation as well as market trends.

New York, New York, Us

Consumer credit portfolio modelling and scoring analytics.During almost two years spent in credit policy department of CitiFinancial I was involved in broad range of activities related to portfolio analytics and modelling. Key areas of my work were:- Portfolio performance and profitability analytics.- Validation of credit policy changes and forecasting the impact on acquisition / performance / expected loss level of credit portfolio.- Optimization of customer segmentation.- Analytical tools design and development (SAS, VB .NET, VBA, Access, Excel)- Portfolio structure forecasting and time evolution (vintage and age structure, delinquency bucket flow effects).- Score model development and validation (application and behavioral).- Scorecard performance monitoring.- Scorecard strategy development and validation.- Credit bureau data processing and analyses.- Preparation of Basel II reporting data.- Ad-hoc analyses reported to the head of risk.

Warsaw, Masovian, Pl

KGHM Polska Miedź is the 6th-largest producer of copper and the 2nd-largest producer of silver in the world. The purpose of this research / engineering project was to create a stand-alone monitoring system of ventilation centre for one of mine shafts. Using the system, vibrational parameters of a set of industrial fans are measured and analysed on real-time basis in frequency / amplitude / time domain. The system controls potential emergency turn-off of the fan caused by increase of vibration level as well as provides full historical data sets for further analysis.The system has been build using National Instrument Real-Time hardware, operating system and development platform.Detailed description of the system has been published in "Diagnostyka" no. 36, 2005 as “Autonomous diagnostic unit for threat identification and risk minimization in technical systems” (by Jedrzej Maczak and Arkadiusz Roszczewski)

Suwon-Si, Gyeonggi-Do, Kr

Digital television software development.As a software engineer at Samsung R&D Center I had an opportunity to work on digital television receiving and processing devices. The role of a device is to tune into MPEG audio/video stream transmitted by satellite or transferred by cable, decompress and generate PAL / NTSC analog signal. This is an essential part of digital TV set-top boxes or TV sets. My key role was related to programming a digital stream demultiplexer chip - a part of chipset based on RISC processor and controlled by real-time operating system. The code has been created in C and assembly.

Austin , Texas, Us

Technical advisory, trainer, conference and exhibition representativeAs an application engineer at National Instruments (global leader in measurement and automation software and hardware) I was involved in variety of customer's projects related to industrial automation and research. My area of expertise was software architecture of LabVIEW development as well as measurement systems. (LabVIEW is graphical programming language and IDE which revolutionized the development of measurement, control and scalable test applications).During the time spent with NI I attended three-month intensive training at the headquarters of the company in Austin, TX. The topics of the training included:LabVIEW programming and application architecture, TestStand - a platform for automated tests, measurement systems architecture and methodology, industrial vision systems, real time software methodology, LabWindows - C based development platform for industrial automation software.

New York, New York, Us

Reporting and analyses of credit portfolio performance and acquisistion. Key responsibiliies of this position were related to design and development of SAS, SQL and VBA based solutions for analysis of consumer credit portfolio. The range of analyses performed includes demographical, behavioral, vintage distributions and ratios as well as forecasting of portfolio quality measures.

Development of corporate credit rating model based on data from Central Statistical Office and Monte Carlo simulations of a portfolio. The main project goal was to prepare and evaluate a set of financial measures of a company in relation to simulated portfolio distributions of financial ratios.

M.S., Computer Physics

Phd, Economics

Neural Networks