In January 2010, I moved my wife and two small boys, ages 2 and 4, to Prague in the Czech Republic to take the role as Chief Operations Officer for Bohemia Interactive Simulations. At that time the company consisted of two offices: a small sales office in Nelson Bay, Australia, and the headquarters in Prague. The two owners, Peter Morrison (CEO) and Mark Dzulko (CTO), hoped to expand the company over the next few years and I was brought on to facilitate that growth.
The company specialized in creation of video game based training and simulation products for military use. With my extensive experience in the same field as Executive Producer for America’s Army, and as Project Manager/ Software Engineer with Northrop Grumman, I was an ideal fit to help bring this fledgling company into the next stage of its development.
My duties in the organization were clear but broad; handle everything that didn’t fall under the existing CEO, CTO and CFO’s purview. This included:
- Information Technology
- Web development
- Customer Support
- Human Resources
- Release Management
- Quality Assurance
The real question for me when I arrived was where to start. The company was a small organization that was ripe for growth, if it were optimized for it. I surveyed the organization and readily identified where the greatest optimization could be made, in the I.T. infrastructure. At that time the company had relied on the sole efforts of Jacob Jarvis, a young but capable I.T. manager. He was completely stretched thin and the software development would only mean more requirements for I.T.. I decided to that I would first focus my efforts primarily in infrastructure to enable growth.
At that time much of the I.T. infrastructure was working but not ideal. The website was on a shared hosting with Dreamhost. The main code line was on a single server, code named “Screamer” because it ran at maximum fan revolutions for some unknown reason, all with minimal disaster recovery precautions. Jacob and I set out to to virtualize the entire server infrastructure as much as possible. We migrated existing services to ESXi 4.x with vCenter. As we did this migration, I steadily setup a new load balancing configuration using HAProxy and Stunnel. The configuration was rather simple, to redirect all traffic to public IP addresses routing in through our firewall (Pfsense) to the HAProxy load balancer. We could then establish multiple Apache and Tomcat VMs in a web farm configuration for that the various web services we would setup. In the end, we had dozens of different services, including SSL, HTTP, Subversion, and FTP routing through load balancers for maximum responsiveness and fault tolerance.
While I made these significant infrastructure improvements, I simultaneously started building functional teams so I could accomplish my longer term objectives. I hired a web development team, I.T. team, and marketing team to begin the process of revamping our customer facing presence. The web site up to that point had not been updated in many years and the company largely was reliant on existing business, with little active development of new sales. I developed a new company and product line branding in effort to re-define the organization on an international basis. From there, I designed and implemented a new marketing site with the new web development team.
Once we had a new corporate branding established, I then started designing a new product licensing system on top of our existing web infrastructure. This new licensing system would serve to greatly simplify the processing of new purchases by customers and reduce the time to delivery from on average 3 days to mere minutes. My focus in all of this was to reduce the overall operational inefficiencies in particular to the sale pipeline process, so as to bolster revenue growth potential. Along those same lines, I would later design and implement a new online quoting system and web store to further push the point of sale conversion to a rapid as possible.
At any one point I generally had at least three large scale projects I was managing simultaneously.
Throughout all these heavily I.T and software development tasks, I also conducted the logistical and marketing planning for large scale conferences, both I/ITSEC and ITEC, as well as relevant marketing material to support them, even to producing and editing promotional videos (link).
Throughout all these activities, I was also busy in other areas of the business, including development of the product line. When I first came on board, the development department had no reliable and repeatable development process. They used a system of emails and a mail reader to manage tasking, which invariably meant lost tasks and untraceable responsibility. I spent time training the CTO on Agile development processes that I had employed while Executive Producer of America’s Army. I also deployed Atlassian products, Jira, Confluence, and Bamboo, in support of these processes. In time the development team used Jira heavily, creating thousands of tasks every quarter and maintained a robust automated build system that used over a dozen servers and hundreds of unit tests.
There was no area of the business that I did not have an active and influential hand in
In early 2012, Peter Morrison approached me with the idea that they wanted to sell the company to a private equity firm. I supported this decision and started the process of creating necessary artifacts for the due diligence phase of the acquisition. By Dec 2012 the process was near completion and the deal had been struck with Riverside. About this time my family and I were starting to feel it was time for us to return to the United States for a time. Though Prague was an interesting place to live, it had it’s challenges as well. We had grown weary with some of those challenges and needed at time of rest and recovery. I resigned from my position as COO in May with the intent to return to the United States.
Soon after my resignation the upper management offered me the position of Chief Information Officer with the open basis of conducting the role from the United States until and effective transition could be finished. I agreed to this position, primarily because at that time I was actively planning the Prague office relocation and knew that the people I had spent time mentoring and training would suffer if I left the project. My family left for the U.S. soon after and I remained in Prague for an additional 3 months to complete the office relocation. Once the project was complete, I rejoined my family in Sandy, Utah and continued the process of training my replacements via telecommuting.
For more pictures of my time at Bohemia Interactive Simulations, check out my Google+ Photo Albums.