Azure AI processes

Every AI-based project is required to go through a certain set of steps before being operational. Let's explore these seven phases:

Data ingestion

In this phase, data is captured from various sources and stored such that it can be consumed in the next phase. The data is cleaned before being stored and any deviations from the norm are disregarded. This is part of the preparation of data. The data could have different velocity, variety, and volume. It can be structured similarly to relational databases, semi-structured like JSON documents, or unstructured like images, Word documents, and so on.

Data transformation

The data ingested is transformed into another format as it might not be consumable in its current format. The data transformation typically includes the cleaning and filtering of data, removing bias from the data, augmenting data by joining it with other datasets, creating additional data from existing data, and more. This is also part of...