The Platform Thesis

Most fintech platforms are using AI to play defense. And calling it innovation. Automating your support, your onboarding, your reconciliation is real. Do it. But be honest about what it is: you are making the thing you already do a little cheaper. That is defense. It protects the base. It does not grow the platform. Growth is the other half, and it is harder. Net-new products and workflows on top of your platform that partners will actually pay for. Here is the trap. If you are a platform, your partners are your distribution. The second you start building the vertical apps they build, you compete with the ecosystem that sells you. So "just build it ourselves" is usually the wrong move, even when you can. Which leaves most platforms stuck: demand for new applications is real and growing, and nobody has the hands to build it. Automating what you already do is easy. Doing more is the whole game. What has your platform actually shipped that is new? #fintech #banking #AI #platforms #productstrategy

"Vibe coding" is a distraction. The real story is the volume. GitHub logged close to 1 billion commits in 2025. Listings for "forward-deployed engineers" jumped roughly 800% in nine months. The cost of turning an idea into working software has collapsed. For a platform, that is not automatically good news. When anyone can generate a rough version of almost anything in a week, feasibility stops being your moat. The bottleneck does not disappear. It moves. It moves from generation to qualification. The question is no longer "can this be built?" Assume it can. The harder questions decide who wins: what deserves to get built, which ideas are worth trusting with real financial data, and who can put a working version in front of a partner fast enough to close. Cheap software did not solve your problem. It moved it. Where did your bottleneck actually go? #fintech #banking #AI #softwareengineering #productstrategy

The most valuable line item on your platform's roadmap is not a model. It is a builder who can sit in the room with a partner. Palantir invented the forward-deployed engineer over a decade ago, for customers who could not tell you what they needed until they saw it running in their own messy environment. AI did not kill that role. It made it the whole game. OpenAI, Anthropic, Ramp, and Cursor are all hiring for it right now. For a banking or fintech platform, that is the missing layer. Not more features. Not better docs. A person who connects what your platform already does to what a specific partner actually needs, and ships a working, trustworthy version before the interest cools. And in banking, speed alone is a liability. Trust is the product. A fast way to ship the wrong thing is just a fast way to lose the account. Separate the rules from the judgment, keep a human in the loop where being wrong is expensive, and make it explainable. The winners will not have the longest feature list. They will have hands. #fintech #banking #AI #forwarddeployed #platforms

New on the Black Flag journal: "Your Fintech Platform Can Build Almost Anything Now. That's the Problem." I spent years building on bank cores at ZSuite and three years shipping production software with AI agents. Here is the pattern I keep seeing on every banking and fintech platform. AI made software cheap to build. That did not remove the bottleneck. It moved it. Feasibility used to separate the winners. Now that anyone can generate a rough version of almost anything in a week, the scarce layer is different: the forward-deployed builder who turns a partner's need into working, trustworthy software fast enough to close the deal. The piece makes the case for forward-deployed engineers in fintech, why "just build it yourself" quietly cannibalizes your partners, and how to decide what is actually worth building. Read it here: https://blackflag.design/blog/fintech-platform-forward-deployed-builders #fintech #banking #AI #forwarddeployed #platforms