The path from building an accurate model in a lab to deploying it in a product involves collaboration of data science and engineering, as shown in the following three steps:
- Data research and hypothesis building involves modeling the problem and executing initial evaluation.
- Solution building and implementation is where your model finds its way into the product flow by rewriting it into more efficient, stable, and scalable code.
- Online evaluation is the last stage where the model is evaluated with live data using A/B testing on business objectives.
This is better illustrated in the following diagram:
