Most enterprise software projects start with a 6-month discovery phase, 18 months of development, and a budget that makes your CFO cry. This one started with a Friday evening and a clear spec in my head from years of watching manufacturing operations struggle with disconnected tools.
I used an AI-assisted council approach — multiple specialised agents working in parallel, each handling a different domain (database models, API routes, UI components, business logic validation). The key wasn’t that AI wrote the code. The key was that I’d spent 25 years understanding what the code needed to do. AI was the force multiplier. Domain knowledge was the foundation.
You don’t need a team of 40 and an 18-month timeline to build enterprise software. You need someone who deeply understands the problem domain, modern tooling that removes boilerplate, and the discipline to ship iteratively. The 48-hour version wasn’t perfect — but it was in production, collecting real data, before most projects would have finished their requirements document.
Start building things. Don’t wait for permission or perfect conditions. The best way to understand enterprise systems is to build one. Use Next.js, Prisma, and TypeScript — they’ll get you from zero to production faster than anything else in 2026. And always start with the data model. If your schema is right, the rest follows.