Accelerating The Value Derived Using AI
A majority of projects with AI at its core are less deterministic. This results in an uncertainty of the value (return on effort) delivered by building a model.
From a technical viewpoint, the uncertainty increases further as there is a dependency with the data engineering team in fetching and making the right data points available, as well as a dependency with the DevOps team to ensure the whole workflow is in production running at the desired SL.
From a business viewpoint, new learnings happen as the data analysis & model building process starts delivering insights – resulting in an even further increase in uncertainty (course corrections in the problem statement).
Two approaches to reducing uncertainty are:
1. Anticipate issues beforehand
2. Accelerate the development process so that one can do more iterations
In this talk, we will touch upon some of the common issues to address and go through multiple accelerators that help accelerate the development process.