Meaningful work

January 11, 2013 10:01 PM

I have a pretty powerful, patented, implementation of machine learning software (genetic programming) and left my company a few months back. I have some time and have been looking at possible jobs at various companies. However, many of them seem trivial to boring, in part because I want to do something more meaningful then analyzing social media data or ad data or (ugh!) financial data. So the question is: What should I do?

A little background: 1. I'm mid-50s and have a good resume, having done some pretty interesting things and been fairly successful but there's nothing in my resume that is like "WOW!"; 2. I am an expert in the use of genetic programming and know a lot about biologically inspired software and a fair amount about machine learning; 3. For various business reasons, working in Internet infrastructure and bioinformatics is not advisable.

Some ideas I have kicked around include working on climate change models/data analysis, alternate energy/smart grid, or more generally, material science. I know a little bit about all of these areas, but not much, however I have a long history of picking up information fairly quickly and finding useful new ways of looking at data or solving problems.

Any ideas for specific problems where a different approach/data analysis might be worth a try? Any specific companies I might approach?

posted by BillW to Work & Money (6 answers total)

I work for Cloudera, a company that helps organizations do large-scale data processing. This mainly involves Hadoop, but a related tool we support is Apache Mahout, an open source machine learning library that can use Hadoop to better scale with certain algorithms. Although many of our customers' use cases are in the same areas you're excluding from consideration (finance, advertisting, Web, biotech, etc.), you might still find some inspiration from the use case articles in our blog.

You might also find Numenta particularly interesting.
posted by tomwheeler at 10:34 PM on January 11, 2013

Check your memail (that envelope in the upper right hand corner of the screen)
posted by infini at 10:34 PM on January 11, 2013

Since you are in Michigan, how about something automotive-related, like traffic data, safety/accident data, etc.?
posted by Dansaman at 11:19 PM on January 11, 2013

The legal industry is fairly enamored with the idea of 'smart' software handling a some of the onerous yet very expensive document review tasks. For example scan all the emails from an opposing corporation and determine a small subset that are relevant to a court case, instead of a small army of rather expensive contract lawyers.
posted by sammyo at 5:15 AM on January 12, 2013

Response by poster: Thanks one & all for comments and suggestions. They are all good suggestions. Some I am sure I am not interested in, some will take some time to evaluate, but regardless of which category your response falls into, thank you all for taking the time to post!
posted by BillW at 2:03 PM on January 13, 2013

Definitely smart grid related things. There is a lot of potential in the industry, a lot of big players interested, a lot of start-ups around, and it all requires big data analysis. Machine learning is extremely useful.
posted by molecicco at 6:20 AM on January 16, 2013