PyData DC 2016
Many organizations still rely on SPSS/SAS to do most of their analytical work. These tools are not only very costly ($10k+ per year per license), but are also limited (no scripting ability, very manual). In 2015, we began transitioning into Python to build robust tools and to reduce operational costs. Along the way, we learned a lot about propagating new tools within a company, reverse engineering, and helping others adjust to a new paradigm. This talk will outline our process of evaluating new tools, initial adoption and companywide propagation. One of the main findings was that open source does not mean free, and we had to take into account each person's experience and comfort in devising our implementation strategy. Lastly, we had to develop our own internal library in order to maintain functionality.