If you work in banking or lending, every software company now claims to be "AI-powered." Your inbox is full of demos and your board is asking questions. Meanwhile, some of the people who built these systems are publicly saying the world should slow down. It's fair to feel excited and uneasy at once.
I want to tell you how Baker Hill thinks about AI, and why I believe you can trust the way we're doing it. The short version:
AI isn't our strategy. Better business lending is.
AI is how we deliver it, inside the workflow, with a banker at the center of every decision. That belief runs through everything we build, which is why I describe Baker Hill as purpose-built for business lending in four ways.
I'm entering my third decade building software for lending, and before that I was a lender. I've decisioned loans, underwritten them, and carried the risk when a portfolio didn't perform. That's why we say we are lending experts enabling AI, not AI experts learning lending.
The difference shows up in the first question I ask about any new technology: so what? What problem does it solve, and does it make sense for a bank to pay for it? A bank CIO I know was recently pitched a conversational AI for his phone system. Each call would cost $2 to $2.50. His existing system handles the same call for about $0.22. In banking, that math matters, because I can't charge my customers an exorbitant rate on a loan or overpay on a deposit.
So we don't put AI everywhere simply because we can. We apply it where it solves a real lending problem. We can do that with confidence because the experience is deep. We've served more than 400 financial institutions, and the average customer has been with us for 14 years. Over the last decade, more than 780,000 commercial and CRE loans have moved through our platform, along with over 420,000 financial statements spread for those loans. That kind of volume comes from 40-plus years of doing one thing.
Business lending is hard, and it runs on documents. Tax returns, personal financial statements, articles of incorporation, covenants. Technology has been trying to read these since the 1990s, and it worked until a document looked a little different, like a personal financial statement formatted differently by every CPA.
This is where AI is already delivering. Our Intelligent Documents & Data captures, classifies, extracts, and validates data from loan documents, and 70 to 80 percent of tax-return-based financial statements now move from manual to automated spreading. The value goes beyond origination. The same capability can track covenants, flood certifications, and renewal requirements across the life of a loan, instead of leaving a scramble for the final 30 days.
Ask BKR gives bankers loan file insights, policy answers, and relationship intelligence without leaving the workflow. We're also developing Opportunity Insights, which analyzes a deal and its financials to bring forward additional perspectives on cash flow and borrower capacity. It's another set of eyes on the numbers, with the banker's judgment at the center.
Because it all runs on one connected platform, you can begin wherever your lending journey starts. A small business loan, a better way to read documents, a faster way to open a business checking account, which is a passion project of mine. I built one of the first cloud-based online account opening products years ago, and I've watched others bring consumer shortcuts to business accounts where they don't fit. Business accounts need the entity verified and then the individual's authority verified, separately. We do that in about two minutes instead of 45. Starting small never means starting over.
You've already invested in the technology you run. We're not asking you to rip it out. Baker Hill connects to every major core, with more than 40 core and technology integrations in all, and it's configurable, so your team can adapt it to your needs without a services engagement for every change.
The same principle applies to AI. Our AI operates inside the systems you already run, not by reaching outside them. It all lives in Baker Hill's private, U.S.-hosted Microsoft Azure environment. Each institution has its own dedicated, single-tenant database, and your data never trains public or third-party models. We're SOC 2 Type II and ISO 27001 certified. Your data stays yours, and so do your decisions.
This pillar is the one closest to my heart: partner, not vendor.
AI makes partnership more important, not less. Every loan is a promise between people, a business betting on itself and a banker betting on them. AI can inform that decision. It shouldn't make it. I'm also clear about where AI shouldn't go. Using it inside the decision itself is not where you want it. Using it to review something and make a recommendation to a human is a safe place to spend our time. At Baker Hill, your institution is always the decision-maker of record.
That's why our belief is simple: AI strengthens judgment. It doesn't replace it. You still have a credit committee, underwriting standards, and a policy to follow. You still need a human who understands whether a loan was underwritten correctly. AI works inside that structure, not around it. Some people imagine running financials through an AI, producing the credit memo, drafting the documents, and closing with one person and no committee. We're not there yet, and we're not going to do that at Baker Hill.
And we build AI to be examined. When your examiner asks how a credit decision was reached, you need an answer, not a black box. Every input is traceable, every model change is versioned, every user action is access-controlled and logged, and a banker reviews the output before it counts. That's a standard your examiner can audit and your board can sign off on, not just a vendor's word for it. The industry is still learning how to govern AI together, and frameworks like ISO 42001 are beginning to appear in master services agreements. I welcome them, because nobody wants to deploy AI in a bank and watch something go wrong.
Partnership also means showing up. Our customer support is 100 percent U.S.-based, out of Indiana. Our implementations are managed and executed by Baker Hill, not handed to system integrators, and every client gets a dedicated project manager and team. A 15-member User Community Board, chosen by their peers from our 400-plus institutions, helps shape where we go next. Twenty-three percent of our clients have been with us for 20 years or more. I've spent the better part of two years visiting close to 400 banks, from CEOs to line-level bankers, and what I hear shapes what we build.
I've seen this pattern before. When the internet arrived, outsiders with no industry knowledge launched startups, and many failed. The technology then moved into established businesses that understood their customers. I expect something similar with AI, only faster.
Smaller loans, say a $250,000 request, may soon be offered and decisioned in seconds, with the back office moving faster than ever. A $10 million commercial real estate deal will still involve real due diligence and human judgment, and it may take years longer to automate. And the relationship stays. Bankers will know more about their customers, with permission, and will spot a cash flow need before the customer asks. You're still going to buy them a cup of coffee.
That's the heart of what we call Intentional AI. Enabling the personal. Take away the manual work so bankers have more time for the conversations that move businesses and communities forward.
If you're thinking about AI in lending, here's what I'd offer:
Start with the problem, and ask what a tool costs per use.
Keep your process. The best AI makes your credit committee and policies faster and better informed, and never routes around them.
Choose partners who know lending. Context matters more than novelty.
Lean in, with guardrails. Competing with national banks means using technology, so don't be the Luddite. Just do it inside a safe, governed environment.
Baker Hill has served lenders for more than 40 years, and we plan to be here for the next 40. We'll earn that the way we always have: by being built for business lending, and by standing behind the bankers who do it.