Hoping to cut parsing overhead and make detection rule writing easier. Currently mapping 20-odd sources.
Any lessons/red flags you can share?