Sema uses machine learning, genetic programming and data analytics to improve software maintenance. Software maintenance is a huge and unpleasant task for all organizations, and I started this business after a year and a half of research, talking to scientists around the globe, based on firsthand experience of the painpoints. In other words, we create tools that help CTOs, developers and architects fix code by understanding and visualizing code quality. We have a team of great technologists- ex Onstar CTO, top professor at U Michigan, DARPA / Bell Labs scientist. Our system can be set up quickly to visualize and recommend solutions to any code base. It can take at least 20% of the time and even more of the risk out of maintaining software.

http://semalab.com/