The world of code is complex. There are multiple programming languages, stacks, architectures, and stakeholders. We could call coding a multi-tenant problem. Anything more complex than
print(“Hello World!”)
is bound to be error prone. The problem amplifies in scale when other people need to read the code.
We are building Stenography to address this issue, and lower the friction of how code is transmitted between humans.
There are great tools out there for coders (Git, IDEs, etc.), but the chasm between the non technical and technical world is vast.
Our hypothesis is that language transformers hold the key. By teaching these language models how to understand both code and human language, we can create the bridge to give code a seat at the table, and allow everyone to participate in the technical growth of a business.
A few code translation examples: