https://craftcx.com - Working on various aspects of trust and observability around the performance of AI Customer Support tooling. How's the support agent performing? Are the resolutions you're billed for "good" resolutions, or just deflecting the customer without helping?