Why I Joined Vectara: A Credit Union Lifer's Take on the AI Moment

I spent the better part of four decades inside the credit union movement. I joined what would become Vantage West Credit Union in 1985 as an accountant who had never set foot in a financial institution, served as its CEO from 2000 through my retirement in 2019, and helped grow it from roughly $500 million in assets to nearly $2 billion, serving most of southern Arizona. Along the way, I had the privilege of chairing the Credit Union Executive Society and Mountain West Credit Union Association and sitting on more industry committees than I can count. I have watched our movement weather the dot-com bust, the Great Recession, the rise of fintech, and a global pandemic.

I share that not to stack credentials, but to underscore the point: I have seen a lot of "this is the moment" moments in this industry. This one is different. That is why, in 2026, I stepped into a formal advisor role with one specific platform, Vectara. This is my attempt to explain why.

What makes credit unions worth fighting for

Credit unions are not banks with a friendlier sticker on the door. We are member-owned, not-for-profit cooperatives. Because credit unions have no shareholders to satisfy, every dollar earned is reinvested in the people who own the institution, the members. It shows up in competitive rates, fewer fees, more lending capacity, and service delivered by someone who knows your name. That isn’t a slogan. It’s the financial architecture of a cooperative.

The result is something megabanks have tried to engineer for decades but can’t manufacture: the trust that naturally forms when your institution is owned by its members, not outside investors. Members of a healthy credit union do not just do business with us, they recommend us, they bring their kids in for first-time auto loans, they call us when their financial life gets complicated. That relationship is the asset. Everything else, cores, branches, mobile apps, exists in service of it.

For two years, I’ve been asking a different question: Are we using technology to strengthen the cooperative bond that makes us who we are, or are we letting it erode the very trust that sets us apart?

Why now, and why "wait and see" is the most expensive option

I have read every breathless headline about generative AI. I am not a hype guy. So let me be precise about what has changed.

A board briefing I have been working through pulls research from Deloitte, Gartner, BCG, and Microsoft. A few numbers stopped me cold:

• BCG estimates a $370 billion retail-banking profit pool unlocked by AI by 2030. Credit unions sit inside that pool, and right now we are watching tier-1 banks and fintechs reach for it first.
• 73 percent of enterprises plan to deploy agentic AI within two years, per Deloitte's 2026 State of AI report. Yet only 1 in 5 have mature governance for autonomous agents.
• Members already expect same-day loan decisions and 24/7 conversational support. "Branch hours" are no longer an acceptable answer.

Here is the part that should keep CEOs up at night. Your members are not comparing you to the credit union down the road anymore. They are comparing you to their bank app, their fintech, and their last Amazon experience. The gap between what they expect and what we deliver is widening, and it is widening faster than most credit unions' technology roadmaps can close it.

A member who waits five days for a loan answer will not wait a sixth. They will find someone faster, and they will not come back. That is the actual cost of "wait and see."

There is also a regulator-side cost. NCUA and CFPB have published guidance on model risk and fair lending. Examiners are already asking pointed questions about agent oversight. A board that cannot demonstrate AI governance is a board carrying an exposure it has not priced.

Start with the business problem. Always.

If I could give every CEO and board chair one piece of advice, it would be this: Don’t buy AI. Buy outcomes that strengthen the member-owner relationship.

Across my career, I’ve watched institutions chase technology for technology’s sake. They build infrastructure before identifying the members’ needs. They buy tools before defining the outcome. And it rarely delivers. Credit unions succeed when they flip the script: they start with a member pain point, a business friction, or a trust gap, then choose the technology that solves it.

For AI in 2026, that means asking three concrete questions:

1. Where are members losing trust today? Long hold times? Slow loan decisions? Inconsistent answers across channels?
2. Where is operational cost compounding faster than revenue? Underwriting capacity? First-call resolution? Document review?
3. Which of these problems would a well-governed AI agent demonstrably move within twelve months?

If a vendor cannot draw a clean line from their pitch to one of those answers, walk out of the room.

The AI Use Cases Poised to Redefine the Credit Union Advantage

After spending the past several months talking with peers, advisors, and the team at Vectara, I have come to believe four use cases give credit unions the most value with the least risk over the next twelve to eighteen months.

Internal knowledge for frontline teams. Your MSRs, loan officers, and back-office staff spend a stunning percentage of the day hunting for information they should have at their fingertips — policy answers, procedure steps, regulator requirements, pricing exceptions. A high-accuracy AI assistant trained on your own documents resolves that on the first contact, every time. This is the lowest-risk, highest-leverage starting point in almost every credit union I have seen.

Member experience and 24/7 support. Members do not stop having questions at 5 p.m. on a Friday. A well-governed conversational agent, one that authenticates silently, resolves the everyday balance-related questions that frustrate members and drain staff time, plastic card fraud, and dispute inquiries end-to-end, and warm-hands off to a human MSR when something needs judgment, is no longer experimental. Early-adopter CUs contain 60 percent or more of Tier-1 servicing inside the agent and lifting NPS at the same time.

Loan processing, end to end. This is the one I get most excited about, because it is closest to our charter. Origination, underwriting, and closing are still wildly manual at most credit unions. A loan officer should not be the assembly line. Early adopters are processing 70% more loans with the same staff, auto-decisioning sub-$25k consumer loans in hours instead of days, getting money to members while the competitor’s banker is still asking for a paystub.

Proactive, agentic outreach. This is where credit unions can finally reclaim the member-relationship advantage. An agent that watches cash flow, anticipates a member's life event, and reaches out with the right offer at the right moment, that is something the cooperative model can do better than a megabank, because we already know our members. We have just never had the tooling to act on what we know.

Why an honest, advisor-led assessment protects the member relationship better than any tool

Here is the uncomfortable truth I have learned from watching credit unions try to "do AI" over the past two years: the technology is rarely what holds them back. The honest self-assessment is.

Every CU I talk to wants to know what AI can do. Far fewer want to honestly answer, "What state are we in to actually use it?" That is a much harder conversation. It involves admitting where the data is messy, where governance is informal, where the talent is thin, where the core system is stuck, and where the organization has not yet aligned on what it wants AI to own versus assist.

I can lead that conversation because I’ve been in the CEO chair for over 20 years, I know the pressures, the tradeoffs, and the mission from the inside. I know the difference between a board agenda item and a board commitment. I know what an examiner will actually ask. And I am only useful in this advisor role if my assessment is honest, even when it slows down a deployment timeline a vendor would prefer to keep.

A meaningful assessment evaluates you across five pillars: strategy, processes, technology and data, culture, and governance, so you know exactly where transformation will stick and where it will stall. It plots you on a strategy-vs-execution matrix. It tells you whether you are a Discoverer (low strategy, low execution), an Operator (strong plumbing, weak direction), a Visionary (bold roadmap, infrastructure behind), or an Achiever (scaling governed agents at pace). Most credit unions under $10 billion in assets sit in the Discoverer or Operator quadrant today. That is not a judgment; it is a starting point.

The point of the assessment is not to grade you. It is to give you and your board a shared, defensible thesis on where agents will own decisions versus assist them, and what has to be true inside the organization for that to be safe.

Why I chose Vectara

I have spent enough time with enough vendors to know how the pitches go. They walk in, tell me they understand credit unions, then show me the same deck they showed a regional bank last week with my logo swapped in. I’m finished being confined to a vertical. Transformation happens when we connect strategy, technology, culture, and member value across the entire enterprise. Credit unions are a mission, and I needed a partner who believed it before they pitched me.

Four things convinced me Vectara is that partner.

1. Obsessive focus on accuracy. When you put an AI agent in front of a member or a regulator, "mostly right" is not a strategy; it's a liability. Vectara is engineered end-to-end for grounded, hallucination-resistant answers, they publish their own open-source hallucination evaluation model and put a Factual Consistency Score on every response. For external, conversational, agentic AI, that is the table-stakes capability most platforms still cannot deliver.

2. Superior multimodal ingestion. Credit unions live in PDFs. Policy manuals, procedure documents, regulator letters, vendor contracts, scanned member docs. Most platforms turn all of that into mush. Vectara ingests text, tables, images, and diagrams as first-class objects, exactly what is needed for compliance-grade knowledge and underwriting-grade decisioning.

3. Out-of-the-box experiences, customizable to each member. I have lived through enough "build it yourself" platform pitches. Vectara ships pre-built interfaces, branded chat, drop-in widgets, search experiences, agent surfaces, that go live in days, not quarters, and sit on top of the same governed platform. That is how a credit union goes from kickoff to live in 90 days instead of 18 months.

4. Managed agents that orchestrate end-to-end work. A loan does not get processed by one model. It gets processed by an orchestrated chain of intake, validation, eligibility, scoring, exception handling, document assembly, e-sign, and post-close QA. Vectara's managed agent platform is built for that orchestration, and that is the difference between deploying a chatbot and reshaping a workflow.

Vectara’s WHY for Credit Unions, and Why I Chose to Stand With You

WHY: Credit unions exist for one reason: to serve people, not shareholders. That belief, that trust, dignity, and community matter, is the reason I agreed to step into this advisor role. Vectara shares that belief. They are not treating credit unions as a vertical to experiment with. They believe our movement deserves the best technology on the planet, and they are committed to making that true.
HOW: They are backing that belief with real investment: engineering resources, productized solutions built specifically for cooperative institutions, and senior advisory time dedicated to our movement. Their leadership has been clear with me: if AI is going to reshape financial services, then credit unions should lead that future, not inherit it.
WHAT: To put action behind that commitment, two offerings are now on the table:
1. An advisor-led AI readiness assessment at no cost.
I am personally leading these for a select group of credit unions. Together, we will: Score where your institution truly stands today, identify the one or two use cases that will move the needle in the next 12 months, produce a board-ready roadmap, whether or not Vectara is ultimately your platform partner, and work through a five-pillar readiness framework
This is about clarity, not sales.
2. A credit union partnership program with meaningful software concessions.
For institutions willing to share their experience and help shape safer, sharper, more useful AI for the entire movement, Vectara is offering significant concessions. This is the cooperative principle applied to technology: the credit unions who help us learn earn the deepest economics, and the entire movement benefits from what we build together.

If This Resonates, Let’s Talk
Whether you’re a Discoverer figuring out where to begin, an Operator with strong plumbing but no strategy, or an Achiever ready to compound momentum, I want to talk. Let’s determine whether your institution is the right fit for the assessment or the partnership.
The agent era is arriving faster than most of us are governing it. Credit unions are uniquely positioned to lead, if we move with the urgency the moment demands and the humility our members deserve.