The Importance of Using a Django Framework to Iterate

When you're iterating, all that we force upon the data scientist is a kind of almost like function framework. You just rather than just writing your query or your pandas transformation or whatever, raw notebook, you just kind of tab it and give it a function name. And then later add versioning and everything's want to productionize it. It will automatically make a very easy productionize while making it feel like you're writing experimentation code. Netflix in this way and the work they have done to make notebooks, like production things right with their Papermill tool. They would literally query the production systems from the training environment.