Search Statement: Robert L

Search Statement: Robert L. Brown

Thank you for visiting my page. This describes my professional experience, a bit of my personal values, and my ideal job.

The thing I do best right now, I believe, is manage software development groups to the goal of releasing products that solve real customer problems. I'm keen on engineering processes, customer focus, professional development (training, learning, mentoring), organizational development (team building), and motivating the group to finish quality products. Currently I am managing, as VP of Product Development, a department of 30 comprising Directors, Managers, and individual contributors, building web-based corporate infrastructure applications in Java. I managed four or five engineering teams at SGI in the late 1990s and, with one exception, released the products. The exception, the FireWalker project, was cancelled and is described below.

My ideal position is a VP or Product Development, managing several software development groups, each having its own Director or Manager, to the goal of creating a product resulting from the integration of each team's component. The technology area can be just about anything, but my recent experiences are in Java/J2EE/JDBC-based applications, 3D graphics (modeling, animation, and internals), authoring systems, and scientific visualization. My approach to development is between "MIL-SPEC" development and "seat of the pants, code it up", is flexible, but is a well stated process. Follow this link to a document that describes my early thinking on the topic: http://www.openeye.com/rlb/swprocess.html. Since 1999, I have been applying the techniques of extreme programming and agile modeling and it is working well.

I prefer a position in a company that needs what I have to offer. This would be a department of good people who are struggling to get products built and out the door. I like solving problems, making groups of people work well together. Give me a real challenge, even some artificial ones, and let me fix it.

I require having good, dedicated, hard-working, honest, smart people around me and have a low tolerance for faulty ethics. I also require that the work I devote myself to has a purpose that aligns with my view of moving civilization and culture forward.

Technically, before becoming a manager I was a prolific programmer. I came to SGI from an institute at NASA Ames Research Center, and was hired in to be the systems architect for the IRIS Explorer project, which was just starting; I was the second engineer hired. In that role, I designed the entire system architecture and wrote the functional spec for most the major components. I was also the lead implementer for the core of the system, focusing on the runtime data typing system and communication subsystem, as well as being responsible for seeing that all of the independently developed components integrated into a single system. One thing I learned going through that exercise is that engineers do not want to throw out things that they've created, yet to integrate the system, this was necessary. Convincing some of them of the importance of this taught me how critical influencing skills are. It also taught me how strong the personal attachment to things one has created can be. In retrospect (at the time) I realized that I had the same roadblock when writing my dissertation; I had to throw out three chapters at one point in order to move forward. That was hard. Peter Denning once wrote a brief CACM editorial entitled "Throw Away Programs" and how they really aren't. It takes discipline.

At the end of the first release cycle at SGI, I became the manager of the group and released four more versions in the next two to three years. This was a bundled product; it shipped with every system SGI sold. Eventually it became very difficult to justify the engineering effort we invested in it, so we sold it to Numerical Algorithms Group in Oxford, England. It is still going strong, and is a favorite system in international high-energy physics circles, such as at CERN.

At Blue Martini, I was promoted to VP of Engineering one month after joining and grew the department to as many as 75 people organically and through an acquisition. The products we built were Java web-based enterprise applications for retail and manufacturing sales-side management. I was personally involved in the engineering infrastructure, product scheduling, development management, and the design of a new major application, resulting in a patent application naming the CEO, the head of Product Management, and me. I also managed the organization through several RIFs, which were well executed.

At Model N, I am the VP of Product Development, reporting to the CEO, and responsible for Engineering, QA, and Publications. Within the first six weeks I restructured the department, put new development processes in place, and had the first major product release out in six months. We have released a product every six months, on the average, since, and added significant new applications.

Other Projects

FireWalker was the second big project at SGI. I took the Explorer team, added a few engineers and a new architect, and began work on interactive title authoring. We had a couple of false starts, including one focused on interactive television, and settled on a game authoring system. I grew the group to about 15 engineers, and we then split it into three groups. I took over the "mastering" group, responsible for conversion and compilation of FireWalker titles on specific targets, such as the PC and Sony Playstation. Despite trips to Sony, Sega, and Nintendo in Japan, we settled on the PC as the first target.

The staff of FireWalker eventually grew to about 24, engineering, SQA, marketing, documentation, and support. After the new architect stepped out, I took over as technical leader over the entire project. However, it became clear in the late summer of 1996 that the return wasn't going to justify the expense, and the ultimate goal of having a repurposable interactive TV authoring system was moot. We disbanded the project, most of the staff departed, and those remaining were transferred to Alias Wavefront.

FireWalker taught me that strong technical leadership needs to be matched with strong marketing leadership; neither alone can drive a successful project. The problem becomes that either area starts to favor their own view of the world over the other's, or even reality. FireWalker was too focused on technical goals, which were ambitious. I find that, even today, I get wrapped up in talking about FireWalker, its technical goals, and architecture. But just as I believe that there is a requirements/spec/design process to follow for the technical side of a project, there is a parallel for the marketing side, which needs to be as clear and as well communicated. If it isn’t there or isn’t getting there, stop everything and do it!

One of the things we did really well is to document the customer process we were fitting into with FireWalker. We devised a story, hired a storyboard artist to draw up the boards, and had a script written to go along with it. After the project ended, I made a movie of the storyboards and voice-over, which we had professionally recorded. It’s amusing, but I still believe that this is a great way to document a customer process/workflow and how your product fits into it.

After FireWalker, I managed a team of engineers (some from FireWalker, some new) inside Alias Wavefront where we designed, built, and released a product, PowerAnimator Worlds, that adds functionality for the games and interactive customers of Alias Wavefront. More recently, we were handed the entire polygon agenda for Maya, AW’s new generation animation system. This work, unlike most in the past, is to be based on NT from the beginning.

After Alias Wavefront consolidated R&D offices, resulting in closing the Mountain View office where I worked, I took up managing four projects in SGI's Advanced Graphics Group, all part of or related to the Fahrenheit project. I have, on the average, a product release every three months for the next year. The first one, OpenGL Volumizer, is already out.

Basic Skills

Software development management, from beginning to end. Some managers specialize in the beginning, some, fewer, in the end. I cover the whole process. I’m a strong believer in doing a lot of up-front work and documenting what was decided, but principally for the purpose of keeping everyone on track with the design and decisions that were made. I have had people tell me "you are the best manager I have ever worked for."

I am a strong software developer. Lately I can claim no specific technology as my forte, but in the past I have wrapped my head around many libraries and methodologies. I particularly like the ideas behind Microsoft’s COM (though the implementation leaves a lot to be desired) dealing with interfaces and aggregation and have developed working knowledge of Java and JSP and use the Tomcat application server for home automation development. I have had people tell me "you are the best engineer I have ever worked with."

These endorsements are very flattering and come from excellent people. On the engineering side, I think it comes from the fact that I have not lost the science of what we do; I have retained a good, practical grip on Computer Science. On management, I believe it comes from being a good listener and being able to explain why we do what we do in terms of the business. Engineers love to hear the "why" of what we do, and meet the "who" that we do it for.

I am a good mentor. It is important to me to teach and learn in whatever I do. I write very well and use regular "news updates" to my department via e-mail. I established a "Manager Discovery" group at Blue Martini Software to teach individual contributors how to think like Managers and I personally run the monthly meetings. I have coached many individuals on their personal and professional development.

I believe in good communication within engineering groups, and in setting up environments were communication is easy and safe. Developing trust is critical. My direct reports feel then can speak frankly and openly with me without fear of repercussions.

I believe in exposing everything that the Product Development department does. We are the biggest cost center in most software product companies and we owe it to everyone to publish a full report on our activities every week. This is very important - I have found that without this, mistrust of the Engineering department starts to creep in.

Some specific things I’ve done in the last few years that are small and unrelated to project management but point to my approach to managing:

Organized and managed a monthly "Manager Discovery" group at Blue Martini. We review topics that face Engineering managers with the purpose of helping individual contributors decide whether or not to move into management.
Organized and lead of co-lead efforts to inject CMM-processes into Blue Martini, to study and integrate pieces of Extreme Programming, and to teach and standardize on the use of UML diagramming in our design process.
Organized and lead a weekly "technology lunch" at Alias Wavefront, MTV. First we studied COM (Rogerson’s book, and these came to be known as the "COM lunches") then Perl then Java. This was open to all of AW’s MTV R&D, but was driven mostly by my group. Efforts like this are hard to keep going, and we would let it relax for spells and then pick it up again.
Helped organized and run a "culture study group" within SGI. We met weekly for two years to talk about cultural patterns and to stir up cultural involvement. The core of the group was four of us; we met with the CEO of SGI (Ed McCracken), presented to his executive team, and presented at an all-exec offsite. This helped lead to a all-management corporate meeting (700 managers!) to create a vision and purpose for the company.
As a part of performance reviews, I give my engineers books. These are not meant to contain specific messages, but to broaden thinking by exposure to new ideas. Some books I’ve given out include Tom Wujec’s "Five Star Mind," Ed Regis’ "Great Mambo Chicken and the Transhuman Condition," Fred Brooks’ "The Mytical Man Month," Bruce Webster’s "The Art of ‘Ware," and David Sheff’s "Game Over."
When my group converted from Unix development to NT development, being faced with a staff knowing nothing of PC systems, I decided (after a fair bit of benefit/risk assessment) to build our own workstations. Basically, one day, we went to the lab, I handed out screwdrivers and boxes of parts, and we spent an afternoon building high-end NT boxes. I have a web page on this entitled "Cuzco NT Workstations.".
Seminar and brown bag organizing; as an effort to bring Alias Wavefront and SGI closer together, I organized several presentations of SGI folks to AW, and vice versa. I ran Brown Bag seminar series at Blue Martini and Model N.

Last updated: 09/15/04
Send comments to Bob Brown