Brendan Farrell

Clockwork Micro

I started Clockwork Micro to offer geospatial microservices that I found a need for while building several geospatial services. We offer software components to make geospatial applications easier to build.

I also do some NLP work. I focus on information retrieval from documents and am always interested in new projects, especially in Seattle.

Recent Years

I left academic research in 2014 to start HowLoud, which offers an environmental noise map of the entire US. From a technical perspective, this was a geospatial computation. In 2016 two companies asked me if I could build a flood data service, and so I did and started National Flood Data. The services that Clockwork Micro offers largely come from pieces of the tech stack necessary for these two companies.

National Flood Data 2016 Flood data for insurance companies
HowLoud 2014 National environmental noise mapping


I grew up in Upstate New York and now live in Seattle with my wife and three children.


I received a PhD in applied math at UC Davis in 2008 and was then a postdoc in the EE departments at TU Berlin and TU Munich and the Computing and Mathematical Sciences departmant at Caltech. I wrote a dozen papers papers in random matrix theory, harmonic analysis and signal processing.

The work most dear to me is work on the discrete or finite dimensional version of Heisenberg's uncertainty principle. Collaborators and I wrote four papers on this topic, but the most basic question (the largest eigenvalue) remains unsolved. I hope to get an email one day with a proof. An article in the Proceedings of the Academy of Sciences describes many areas where this conjecture likely holds, but the simplest and most beautiful, and possibly most difficult, is the Fourier submatrix.

Glory Days

I love connecting again with old friends, neighbors, colleagues and classmates. If we know each other from St. Clement's Elementary, Sarotoga Springs High School, Congress Bundestag Program, Uni Tübingen, Cornell, McGill, UC Davis, TU Berlin, TU München or CalTech ....drop me a line.