What do we mean by a first principles build?
Sometimes the data science problem you face can be completely unique and as such there are no tools or techniques out there than can be applied directly to your problem. In such cases we need go back to first principles and build you something from scratch.
This is not as daunting as it sounds. We start with a simple prototype and gradually add complexity until we have something that meets your needs.
Where have we used it?
We have built software that detects driving events from phone sensors; extracts the financial knowledge from market experts; simulates the spread of infections in hospitals; and feasibility tests biotech products.