When a candidate attempts a multiple choice question, the answer is evaluated by Ada. If the candidate picks an incorrect option, Ada understands what they were missing in their understanding of that concept, which is why the candidate picked that particular incorrect option. Ada can give the candidate a hint explaining the concept in brief, and give them another shot. This provides a friendly interview like experience for the candidate and they might learn something new. Also, it enables Ada to score candidates granularly. Ada can provide increasingly more revealing hints to distinguish between candidates who know the concept, are familiar and can pick up on the job, not familiar with the concept but know the basics enough to pick up the skill fast enough on the job versus candidates who are not familiar with the basics either.
An important distinction is the kind of MCQ questions Ada asks. Unlike the status quo of judging a Java programmer's knowledge based on whether they know who invented Java, Ada asks questions that measures on-the-job skills.
Ada can randomize questions, the options for each question and have multiple variations of the same assessment, so candidates find it harder to use unfair means.