Most reps go into their first discovery call of the week having practiced exactly nothing. Not because they don't care, but because traditional roleplay is time-consuming, awkward and entirely dependent on a manager having a free thirty minutes. So it doesn't happen, and reps learn on live deals instead. AI roleplay changes that equation entirely — and in this session, Sriharsha Guduguntla, CEO of Hyperbound, sits down with Stuart Kerst, Head of GTM Productivity at Klaviyo, to talk about what that actually looks like at scale. Stuart will share how Klaviyo uses Hyperbound to ramp new reps faster, prepare teams ahead of product launches, and keep skills sharp through continuous training — so that objection handling, discovery questioning and late-stage negotiation get repped on demand, with immediate feedback, before they ever show up on a live deal. Together they'll dig into what good really looks like: how to compress time to competency, surface skill gaps before they hit the pipeline, and build the kind of rep confidence that only comes from genuine repetition. They'll also get honest about where the technology still has limits, and how enablement teams integrate AI roleplay into a broader readiness model without it becoming just another tool nobody opens after the first month. If your reps are still learning on the job, this conversation will show you there's a better way to prepare them for it.