BusinessNovember 7, 2026
AI can build the app. Who builds the business around it?
Why Aurekai treats software, engineering and operating knowledge as assets shared across businesses.
NGNick GonzalesNovember 7, 2026 · 3 min read
A working application used to take a great deal of engineering time to produce. That hasn't vanished, but AI has changed how much a small team can build. What hasn't changed nearly as quickly is everything around the application: who needs it, how their company works, where the data lives, what they're willing to pay, and who keeps the result running six months later.
That tension kept coming up in our Aurekai planning conversations. We could imagine an agent building a new product over a weekend, then spending weeks trying to connect it to accounts, policies, customers, support, invoices and the rest of a business. Producing the code was increasingly the easiest part.
The unit of work is getting bigger
Aurekai is our attempt to work on the whole business at once. An independent store might need a website, inventory, customer accounts, invoices and staff tools. A publisher might need profiles, recordings, subscriptions, distribution and community pages. They can look very different to customers while sharing much of the same underlying engineering.
That is why we talk about engineering as capital. When we build a useful catalogue tool, it can serve more than one retailer. A source-library component can help a publisher, a research workspace or a creator. An identity and permissions system can support both of them. The next opportunity can begin with something substantial already in place.
A small team still needs deep expertise
AI helps with unfamiliar code, but operating a business still requires judgment about tax, finance, privacy, fulfillment, labor, customer relationships and service. We don't want those disciplines replaced by a generic prompt. We want applications and tools that make the expertise available at the point where it's needed.
Bonfÿre is the native platform we use to organize that work: Matrix for people and sources, WorkGraph for tasks and capabilities, application grammars for CRM and accounting, model tools for analysis, and public web systems where people interact with the results.
The ambition is a company that gets better at building and operating companies. Its public products are where people encounter the work. The shared software is what lets the work accumulate.
For a fuller explanation of the business model, read Aurekai and Bonfÿre. The U.S. Chamber's 2026 report is also useful context for how many technology systems small businesses now manage.
| Work layer | What automation contributes | What still needs an owner |
|---|---|---|
| Product | Implementation and technical iteration | Customer fit, acceptance and service |
| Operations | Shared records and repeatable workflows | Fulfillment, staffing and accountability |
| Finance | Forecasts and connected calculations | Approval, billing and settlement |
| Continuity | Tests and deployment candidates | Live readback and ongoing support |
Engineering capacity
Aurekai treats developer effort and software as reusable investment across different customer businesses.
Operating expertise
Tax, privacy, finance, customer relations and fulfillment retain specialized owners; a generic prompt does not replace their judgment.