For those running AI voice calling agents (Retell/VAPI/Bland etc.) for outbound lead calls — what are the actual compliance rules you're operating under? Specifically: 1. Consent/disclosure for AI-generated calls (TCPA if calling into the US, DND/TRAI registry if calling into India) 2. Call recording disclosure requirements 3. DNC list handling and opt-out flows And on the build side — how did you actually implement this to stay safe? Is it baked into the voice agent's script itself (disclosure line at call start, opt-out keyword detection cutting the call), or handled as a separate pre-call layer (DNC check + consent logging before dialing)? Would love to hear what's actually worked for people running this at scale.