Congrats to the MindX Service AI team — I think the idea of agentic AI that actually resolves issues (not just replies) is really exciting. Do you have benchmarks for intent detection / resolution accuracy as volume scales (say 1000s+ tickets)?