Implementing New Ideas in a Large Organization

This rotation at the AMSTC in Michigan for GE Capital has been to lead an innovation project. The CIO at GE Capital started an annual ‘Innovation Challenge’ where employees submit innovative ideas they feel can help improve our internal or external processes. The ideas are vetted in several ways and the top teams are invited to present to a panel of judges made up of Senior IT Leaders. The three ideas that win are given a chance to be proven out. This project is one of the top three ideas chosen from the 2011 Innovation Challenge.

The premise of the idea was to create a way to discover ‘components’ being developed by the different teams across the organization. His vision was a web portal through which developers, architects, and others could search and find components for re-use. While it sounds simple and great, the implementation of an idea is complex and presents many challenges. Since my assignment leader has his usual day to day responsibilities, an IMLP was assigned to lead and carry the project forward.

The first part of the project was like being a consultant. A business case had to be built. This involved gathering data on projects in the organization and estimating how we might measure potential efficiencies gained and finding the value proposition for the project. After that the case had to be shopped around to everyone, then reformed based on the feedback received. It felt very entrepreneurial.

Simultaneous to presenting and building support and interest work began on the details of how to actually implement the idea.  Some of the biggest challenges have been hashing out the details and mechanics. Some of the big decision points have included:

• How do you govern what is in the searchable environment and ensure that it isn’t creating more problems?
• Where will the solution fit in the current environment and team structures?
• Who/How will it be supported?
• Who should be able to have access to the tool?
• What risks are we exposing the organization and projects to by opening their code to other developers in the company?

None of the answers have come quickly, in fact, at every turn there is another opinion. Howver, while some questions and challenges still remain unanswered great progress has been made. The scope is to start at a code level. The solution implemented will have a search capability across the diverse repository environments currently at GE Capital. Several products were analyzed and one chosen that aligned with the goals and financially made sense.  It will allow us to hook to several repositories using SVN, Git, and CVS, it also works with specific SCM tools like Borland StarTeam, MS Visual Source Save. The process is simple; the applications will check out code then indexes it in a large database. A user will be able to search, like on google code search, on the projects indexed via a web interface. We are also exploring the possibility of integration with some IDE's.

 The Second part of the project has been the more technical piece, setting up the pilot environment, including provisioning servers, installing software, and configuring the systems. Building out the infrastructure has its challenges as well. Know how to trouble shoot when things don’t run or install the way you expected, who to talk to, and where to go to request services you can’t perform. This stage has included utilizing the network from the first part to the project to build a team to participate in the pilot. As the New Year approaches the in depth evaluation with the pilot team will be the ultimate platform to prove out the concept and benefits proposed.

If all goes well, by the time rotations wrap up, what was only an idea five months ago will be a proven business case ready to be implemented.

What ideas and plans do you have for 2012?