The Fini Podcast

Rohit Tripathi has spent 10 years in technical support at Microsoft, HP, Zapier, Notion, and Automate.io. At Zapier, he was part of a two-person pilot team supporting their entire AI product line while tackling a 700-ticket backlog. In this episode, he breaks down the biggest challenge of supporting AI products in production, why the feedback loop between support and engineering determines success, and what hasn't changed in technical support over the last decade. Plus: his strategy for triaging bugs vs. hallucinations, the one thing he'd automate first, and why the best support engineers aren't the ones who know the most.

Maisa Breve spent five years at Royal Caribbean before leading customer experience at Vio.com, a travel platform serving 100 million visitors annually. In this episode, she breaks down why building trust digitally is harder than face-to-face service, how Vio.com maintains a 10-minute response time for critical issues, and why AI will never replace human judgment for travel exceptions. Plus: the one ticket type AI can't handle yet, and why she believes empathy can't be trained.

Tamara Wall has led support at Common Room, Culture Amp, and Datto. Her take: AI doesn't fix broken support orgs, it exposes them. In this episode, she breaks down what it actually looks like to onboard AI like an employee, why CSAT is an outdated metric, where she draws her red lines on automation, and how the entry-level support job is about to change completely.

Eli Winderbaum has been running an AI support agent for nearly two years at Mirage. In this conversation he shares what actually changed: which roles disappeared, why stale documentation kills AI accuracy, how a customer bug report can go from ticket to fix in under an hour, and why the rest of the world is way further behind on AI adoption than the tech bubble thinks.

Zack Hamilton has spent over two decades advising more than 200 retail brands including Macy's, Sephora, Home Depot, and Coach. His message is blunt: most CX programs get defunded not because they fail, but because they become irrelevant to the business.In this episode, Zack breaks down why NPS creates dangerous false confidence, how to build a one-page business case that actually gets funded, and why most AI in customer support is just "automation of mediocrity": solving the same tickets over and over without fixing the upstream friction.His advice for CX leaders: stop managing dashboards and start driving decision velocity.

Christopher Lind has turned down countless Chief AI Officer roles. Not because he doesn't believe in AI — but because most companies are setting these roles up for catastrophic failure.After 20+ years leading digital transformation at GE Healthcare, ChenMed, and AbbVie, Christopher breaks down what separates AI implementations that work from the ones that quietly drain budgets. We dig into the "AI hangover," why speed is the most overrated metric in CX, and how to spot "work slop" — AI that creates the illusion of progress while building nothing real.His advice for CX leaders planning their 2026 strategy: "If it sounds too good to be true, it is.