Three production SaaS apps built in 48 hours

The Challenge

A recruitment industry partner had a meeting scheduled with an interested company on Thursday afternoon. They needed to demonstrate real, working AI tools — not mockups or pitch decks. The ask: show that AI can genuinely help professional services businesses with document review, client reporting, and customer communication.

The timeline was non-negotiable. Two days to go from concept to three deployed, paying-ready SaaS applications.

The Solution

We built three standalone SaaS applications, each solving a real business problem:

Tech Stack

The Overnight Build Advantage

Speed came from smart architecture decisions, not cutting corners:

• ▸ Shared payment gateway — one Stripe integration serves all three apps. Buy tokens once, they credit across the platform.
• ▸ Token-based billing — simple usage model (Starter 50/$19, Pro 200/$59, Business 500/$119 AUD). No subscription complexity for MVP.
• ▸ Consistent stack — FastAPI + SQLite + vanilla JS across all three. One pattern, three products.
• ▸ Containerised from day one — each app runs in its own Docker container with Cloudflare Tunnel to a custom domain. Zero-downtime deploys.
• ▸ AI does the heavy lifting — Claude handles document analysis, report generation, and conversation. The apps are well-engineered wrappers around genuine AI capability.

The Outcome

All three applications were live and demo-ready before the Thursday meeting. The client saw real products — not wireframes, not pitch decks. They could log in, upload a document, generate a report, and chat with a trained AI assistant.

48-Hour Timeline

Key Takeaways