I am working on a project right now that I just had to scrap a few weeks of work because I realized the agent was agreeing with me when it should have been saying the testing was wrong. It found positive results in many failed tests. I have had to start over with rules in place and constantly reference them when getting results to ensure it is giving me the right overview. Then I still manually validate what it shows.
So far it has helped a bunch. Many ideas I have had, have been proven wrong and alternative ideas have come up in their place. It has resulted it much better progress and improvements over the original path I was going.
So far it has helped a bunch. Many ideas I have had, have been proven wrong and alternative ideas have come up in their place. It has resulted it much better progress and improvements over the original path I was going.