Resume

• Scientific software engineer with an emphasis on C++ and finance
• Proven track record of delivering high-end commercial software products
• Respected author, speaker, and contributor in the C++ community
• Former research mathematician in algebra, logic and symbolic computation
• Experience in a wide range of languages including various Lisp dialects, Unix shellscripting, Perl, Java, C#, C, C++, as well as Web development using HTML, css, and javascript.
• Commercial development experience mostly in the Windows environment, but well-versed in the "Legg" environment (Linux/emacs/gcc/gdb) as well

The company I currently work for licences software for performance and style analysis of stock and bond portfolios to leading institutional investors, money managers, and financial consultants worldwide.

Prior to 1997, the company had been marketing a 16-bit windows program that had been written by contractors. In the fall of 1997 the company embarked on a complete rewrite and extension of its software in C++ for Windows NT, using its own team of staff engineers. I was the second person to be hired for the project.

My first and main responsibility within this project was the design and implementation of a backend calculation engine in C++, using design and implementation technologies such as design patterns, multithreading and STL.

At the time, design patterns were just becoming widely known and accepted. The acceptance level of the STL in applied software engineering was practically zero. It may not be an exaggeration to say that ours was the first large commercial software project that used generic programming and the STL throughout. Needless to say, we also became early adopters and heavy users of the Boost library.

Similarly, the way I used multithreading extensively in the math engine was rather ahead of its time: its immediate purpose was to enhance the user experience of the program by deferring lengthy calculations to background threads. However, the design was such that it can now be leveraged to obtain better performance on multiprocessor machines.

Implementing the math engine and the multithreaded calculation framework gave me ample opportunity to hone my unit testing skills. At one point, about one third of the way through the project, my unit test collection had more lines of code than the entire rest of the project.

The mathematics of the engine that I wrote is based on Modern Portfolio Theory. In particular, the module implements returns-based style analysis as set forth by Nobel laureate and Stanford professor William F. Sharpe.

Credit for the overall design and architecture of the software goes to our CTO/project lead and not to me. However, since our team consisted of just the two of us for several months, I was always passively and sometimes actively involved in the high level design decisions.

As the team grew from 2 to 5 developers and the project matured, my work broadened beyond the application’s back end. Firstly, there was significant collaboration with other team members on the design and implementation of the middle layer between the mathematical engine and the user interface part of the program. Later, I also performed significant work on the presentation layer, including the charting engine.

The product was shipped in December of 1999 and is now being used by more than 600 plan sponsors, money managers and financial consultants worldwide.

In the spring of 2000, the company began development of a Web-based version of the software in a team consisting of two to four people, at various times. My role in this team consisted of the following:

• Participation in the overall design of a distributed Web server architecture, involving a load balancer running under Linux and backend servers under Windows NT.
• Development in C++ under Windows NT for the backend servers.
• Java Servlet development under Linux, including multithreaded programming and socket programming.

The completion of the first version of the Web product marked the end of the startup phase of the company. At that point in time, it would have been natural for me to seek new challenges. An important reason for my decision to stay on was the fact that I had a financial stake in the company, and the outlook was bright, not in the least because of the quality of my work. From hindsight, the decision to stay has turned out to be more than just financially advantageous. It has also provided me with the unique opportunity to experience the growing and maintaining of a large code base over an extended period of time. Conceiving of a piece of software, building it, and getting it to market in due time is one thing. Staying with it for years, seeing how early design decisions and implementation techniques stand the test of time, is quite another.

Another aspect of my work which has done a lot for my maturity as a software engineer is the fact that here, the feedback loop between the engineer and the end user is extremely short. Nothing gives a software engineer a better reality check than to work for a small company with a dedicated and demanding community of users who will let you know what they think and who will abandon you for the competition if you cannot deliver quality in a timely fashion.

The growing client base of the company has also created a role for me as staff mathematician. Besides working on the software, which still takes the majority of my time, I also tend to all things mathematical, including client support, research concerning ongoing developments in portfolio theory, and talks at user conferences.

Finally, I also serve as a mentor and instructor of more junior developers on the team, primarily on the subject of C++ and the Boost library.

Topics of the training classes that I taught for IT professionals were: Programming in C++, Windows NT Programming (Win32, MFC, COM, ODBC), as well as Windows NT System Administration and Networking.

• Research in computer algebra (see list of publications)
• Design and implementation of mathematical algorithms in the computer algebra system Reduce
• Teaching undergraduate and graduate level mathematics courses with an emphasis on computer algebra
• Co-authorship of a graduate level textbook in computer algebra (see list of publications)