In the latest episode of the Leadership in Insurance Podcast, host Alex Bond sits down with Kasey Roh, Head of US at Upstage AI, for a conversation that stretches from the lessons of large-scale manufacturing automation to the cutting edge of AI-powered document processing in insurance.
Kasey brings a rare perspective โ shaped by hyper-growth stints at Meta and Tesla โ to the world of enterprise AI.
She reflects on witnessing Tesla's journey to scale Model 3 production, where the hard-won lesson was clear: attempting to automate before truly understanding every step of the process is a costly mistake. Only by mapping workflows end-to-end with domain experts first did the operation find its footing โ a principle she carries directly into how Upstage approaches AI adoption in insurance today.
In this episode, Kasey and Alex explore:
๐น Why understanding end-to-end workflows is non-negotiable before any automation begins
๐น How Upstage made the leap from the Korean enterprise market โ where it now powers over 60โ70% of claims processing โ to cracking the US market
๐น The critical difference between how Asian and North American enterprises buy technology
๐น Why OCR and LLM technology is transforming underwriting and claims processing, handling the complex "tail end" of documents that traditional systems simply can't
๐น Whether AI will truly disrupt existing carriers, or whether it's the new AI-native players who will reshape the industry first
Kasey also shares her candid take on AI adoption in insurance: the proof-of-concept phase is ending, and organisations that haven't figured out governance and human oversight are going to struggle in the scale-up.
Whether you're in underwriting, claims, or leading digital transformation at a carrier, this episode is packed with hard-won lessons from someone who has scaled technology at some of the world's most demanding and successful companies.
Hosted on Acast. See acast.com/privacy for more information.