Background
Fermentation is as much of an art as it is science. In nearly all fermentation processes the challenge is to create the right environment with the right food for the right microbial strain to yield the desired product.
Timing, temperature, feed rate, pH, and many other factors have to oftentimes be in fairly narrow ranges in order to yield the optimal result. Additionally these ranges are rarely known ahead of time and must be discovered through a combination of different methods, such as trial and error, intuition, statistical analysis, and now machine learning.
The latter two methods require the use of computational resources which means the data around these fermentation processes must be able to be parsed and understood by a program. However the first two methods are largely human driven which means we need language that can model the fermentation systems and processes and be both parsed and understood by machines and read and understood by people.