Automating Data Collection For AI At Morningstar

One of the biggest challenges in making AI projects a success is dealing with the requirements for data needed by machine learning systems. Machine learning systems work by generalizing learnings from data, so if that data is insufficient in quantity or poor in quality, then the machine learning project will fail. Nothing is more true for artificial intelligence than the tech adage, "garbage in is garbage out". Shariq Ahmad, head of technology in the data collection group at financial services data firm Morningstar knows this very well. As part of his role at Morningstar, he is responsible for building a pipeline and methodology for dealing with large quantities of data in a wide variety of formats, qualities, and levels of completeness and accuracy to support big data projects, including those that support their machine learning efforts.