Building an AI Startup in 2026 (the last 3 months changed everything)

What Lawyers See Before Credit Union AI Deals Go Wrong

Ankur Patel interviews Michael Heller, Associate Attorney at Messick Lauer & Smith. 📩 Subscribe to the Main Street AI Newsletter: https://multimodal.beehiiv.com/subscribe?utm_source=podcast&utm_medium=description&utm_campaign=podcast Michael Heller is an Associate Attorney at Messick Lauer & Smith, where he focuses on vendor management, technology contracts, due diligence, and fintech partnerships for credit unions and CUSOs. Guest: Michael Heller, Associate Attorney, Messick Lauer & Smith Host: Ankur Patel, CEO & Founder, Multimodal 0:00 Introduction 0:42 How Michael started in vendor management for CUs 2:56 Credit unions, CUSOs, and where fintechs fit in 4:14 The nature of the deals: vendor, M&A, and CUSO formation 5:32 Where credit union and tech vendor deals actually go wrong 7:39 Regulated stability vs. move-fast disruption 9:26 Honeymooning and the seeds of discontent 12:16 Why agentic AI deals are harder than traditional software 14:53 Clear exit pathways and a robust change management process 16:32 Indemnification, liability caps, audit rights 17:58 Building a risk framework 20:00 Tricks of the trade 24:13 Setting up a long-term partnership and allocating risk 25:00 AI in credit decisioning: who holds the liability? 26:13 You can't outsource risk 27:44 AI as a tool: the human in the loop stays on the hook 29:44 State vs. federal AI regulation 31:16 Right-sizing AI adoption for smaller credit unions 34:17 How AI is reshaping the legal profession 37:08 Why lawyers are a shield and the cost of getting it wrong 38:14 What a credit union that gets this right looks like in 2-3 years 40:08 Closing

Why This CU Rebuilt Its Entire Digital Front Door

Ankur Patel interviews Courtney Rowan, SVP and Chief Digital and Transformation Officer at Citadel Credit Union. Courtney has spent 27 years watching digital go from a side project to the only front door, and she brings the rare full-arc operator perspective on what it takes to make that shift work. She breaks down the non-documentary onboarding rebuild that lifted funnel conversion 212%, the CDP/CXP stack overhaul behind true member individualization, and the three-pillar AI steering committee setting Citadel's strategy. 📩 Subscribe to the Main Street AI Newsletter: https://multimodal.beehiiv.com/subscribe?utm_source=podcast&utm_medium=description&utm_campaign=podcast Guest: Courtney Rowan, SVP and Chief Digital and Transformation Officer, Citadel Credit Union Host: Ankur Patel, CEO & Founder, Multimodal 0:00 Introduction: Meet Courtney 0:55 What Courtney owns at Citadel 1:54 Digital evolution: from side project to the front door 5:00 Rewiring piecemeal digital channels so AI can be excellent 7:51 One continuous experience across every channel 10:04 What a charter expansion actually involves 13:07 Onboarding with no branch: human-led, digitally enabled 18:07 The fintech onboarding playbook 19:46 The results: 212% funnel conversion and more 21:35 Gen Z, values, and why credit unions get stuck 25:15 Lots of data, no unified brain: building a CDP and CXP 28:00 Don't drive a Porsche like a Ford 31:13 Standing up an AI innovation steering committee 32:52 Governance, education, enablement 36:18 Adopting AI individual by individual 37:57 AI as copilot, not a layoff plan 39:02 Separating AI marketing hype from reality 41:03 How Citadel employees use AI today 44:15 Working with incumbent and new AI vendors 46:58 The 3-5 year vision: banking that feels effortless 48:55 Closing

How LMCU Cut a 150-Step Process to 15 with AI

Ankur Patel interviews Chris Ortega, who leads technology transformation at Lake Michigan Credit Union, a $17B asset institution and the 13th largest credit union in the US. What's distinct about Chris's approach is that he restructures how the organization operates before deploying the tools — agile teams, a Center of Excellence, and a single Emerging Technology Governance Board that owns AI, RPA, and stablecoin. The episode covers how that structure was built, where RPA vs. agentic AI is the right call, and the cost discipline that keeps emerging tech from running away from the institution. 📩 Subscribe to the Main Street AI Newsletter: https://multimodal.beehiiv.com/subscribe?utm_source=podcast&utm_medium=description&utm_campaign=podcast Guest: Chris Ortega, CIO, Lake Michigan Credit Union Host: Ankur Patel, CEO & Founder, Multimodal 0:00 Introduction: Meet Chris Ortega 0:32 Stabilizing the IT landscape at LMCU 1:40 Inside a top credit union 3:23 Building conviction for transformation 5:20 The Center of Excellence and the three C's 7:43 What surprised him about emerging tech 9:31 Where traditional project management falls short 11:25 Why agile fits today's pace 14:06 One committee for AI, stablecoin, and whatever comes next 15:53 From planning to execution to ROI 18:33 Closing the gap between business and engineering 22:09 Building trust through real back-and-forth 24:35 Why remote teams are an asset, done right 26:44 What smaller credit unions can take away 29:20 Pick your tech, then match your partners to it 33:25 Where agentic AI is showing up 35:45 The HUMDA win: 150 steps cut to 15 36:56 Mortgage, the call center, fraud, and audit 38:51 The token-bill trap 40:46 RPA vs. agentic AI 44:13 AI bootcamps, lunch and learns, and Co-Pilot 48:42 Star Trek, not Terminator

UKFCU's Change Agenda: New Mortgage Dept, Zest AI, First Merger

Ankur Patel interviews Lisa Highley, Chief Lending Officer at University of Kentucky Federal Credit Union (UKFCU). They chat about the credit union's new mortgage department, a rebuilt commercial lending operation, Zest AI live in consumer lending, GLIA live in the call center, and the institution's first merger. 📩 Subscribe to the Main Street AI Newsletter: https://multimodal.beehiiv.com/subscribe?utm_source=podcast&utm_medium=description&utm_campaign=podcast Guest: Lisa Highley, Chief Lending Officer, University of Kentucky FCU Host: Ankur Patel, CEO & Founder, Multimodal 0:00 Introduction: Meet Lisa Highley 1:03 UKFCU at $1.75B and the full lending portfolio 2:30 Why the change agenda 4:01 Diversifying out of indirect auto and home equity dependence 5:46 Branch referrals driving the new mortgage department 7:05 Build vs. buy 9:04 What Empower automates in mortgage origination 13:02 Document review pain still left for AI to solve 17:13 How many hours a mortgage file still takes today 18:27 Staying competitive in indirect auto 22:00 Where software actually helps 24:52 Why same-day funding matters to dealers 25:57 Zest AI rollout, year-one goals, six-month review 29:04 How UKFCU decides what Zest can auto-approve 32:11 GLIA on day two 33:30 Reducing after-hours call support by 70% 34:30 Voluntary staff repurposing, not layoffs 35:43 The case for AI close to the member 37:03 Lessons from UKFCU's first merger 40:17 Two-to-three-year vision

How Capital Credit Union Uses AI in Marketing: From Strategy to Adoption

Ankur talks with Steve Zich, Chief Marketing Officer at Capital Credit Union in Green Bay, Wisconsin, about an operator's playbook on deploying AI in the marketing department. The conversation goes deep on personalization done right (vs. creepy), competing with big banks and fintechs on community rather than spend, and the team-level mindset shift that makes AI adoption stick. 📩 Subscribe to the Main Street AI Newsletter: https://multimodal.beehiiv.com/subscribe?utm_source=podcast&utm_medium=description&utm_campaign=podcast Guest: Steve Zich, Chief Marketing Officer, Capital Credit Union Host: Ankur Patel, CEO & Founder, Multimodal 00:00 Intro 00:38 What a credit union marketing operation actually owns 01:34 Filtering AI signal from marketing theater — Steve's three filters 05:06 Differentiating brand voice in a world driven by AI 07:56 Rolling out Jasper 12:56 Measuring ROI 14:29 Content cannibalization 16:45 What good personalization actually feels like 21:49 Personal vs. creepy — drawing the line in member outreach 24:34 Competing with big banks and fintechs on community + data 33:17 The fear narrative around AI 38:57 Why generations are shrinking 40:23 Closing