Front End == Back End?

Typically, software developers have been grouped into one of two groups: chino wearing, “how did you get your style?” front-end and f*ck-you flip-flop wearing, “how did you figure that out?” back-end. However, the rise of tech industry buzzwords such as “machine-learning”, “big-data”, and of course “the-internet-of-things” bely a kernel of truth: these new technologies are altering the development landscape and even alleged “full-stack” developers’ roles are changing.

Increasingly, software and products are relying on the “back-er-end” developers. These developers, generally termed “data-scientists” who actually write the code and conduct the analyses, typically have little if any front-end experience. Furthermore, even technically proficient front-end developers rarely understand the details and assumptions underpinning these algorithms. Historically, this did not matter. The algorithms just needed to work. However, with the proliferation of data and the increasing utilization of it to automatically make impactful decisions, deeper understanding is key.

The front-end design of data-visualization is now directly tied to the back-end algorithms making decisions on N-dimensional data. Users will want to see how an algorithm is making its decisions. People want to know if, how, and why their Roomba, Nest, or self-driving car ran amok.

The challenge of the front-end is that they need to help humans see the “thought-process” of the algorithm. Data-visualizations must now distill N-dimensional computation into a two or three dimensional model that a non-math major can intuitively understand. The challenge of the backend is that they now need to not only build the data-flows but also help guide the design process to make the visualizations not only beautiful but also compelling representations of reality.

If “big-data” and “machine-learning” hope to live up to their hype they will require ever deeper collaboration and understanding of both the technical back-end and front-end design processes. I am not sure what this will do to the ratio of chinos to flip-flops in the average data-driven organization but I do know that the age-of-automation is upon us and it is here to stay.