Crew Craft is a workforce scheduling platform for hospitality, events, and retail. It handles scheduling, job postings, and applicant tracking — from vacancy to first shift. Over 10,000 shifts have been planned through the system.
But this article isn't really about workforce scheduling. It's about how the platform was built, and what that says about the future of software development.
AI Pair Programming
Crew Craft was developed using a pair programming approach borrowed from Extreme Programming (XP). XP is a disciplined Agile software development framework designed to improve software quality and responsiveness to changing requirements through high-frequency, short-cycle iterations. It emphasises technical excellence — practices like pair programming, test-driven development, and continuous integration — to create sustainable, high-quality code.
In traditional pair programming, two developers work together: one writes code (the driver), while the other reviews and guides (the navigator). They switch roles regularly. The constant dialogue catches errors early, spreads knowledge across the team, and produces better-designed code than either developer would write alone.
I've adapted this model for AI-assisted development. I take the navigator role — setting direction, making architectural decisions, reviewing output, and course-correcting when needed. The AI takes the driver role — writing code, implementing features, and handling the mechanical aspects of development.
This isn't about prompting a chatbot to generate snippets. It's a continuous collaborative session where context builds over time, decisions compound, and a real product emerges from the partnership.
Why This Works
The navigator/driver split plays to each party's strengths:
- Human strengths — Domain knowledge, business context, user empathy, architectural vision, quality judgement
- AI strengths — Speed of execution, broad technical knowledge, consistency, tireless iteration, pattern implementation
I bring years of experience in workforce management, understanding of how festival catering operations actually work, and the ability to recognise when something doesn't feel right. The AI brings the ability to translate that understanding into working code, fast.
The result is a development velocity that wouldn't be possible alone, combined with a coherence that wouldn't emerge from AI working unsupervised.
The Investment
Crew Craft is backed by three investors who understood the potential:
- Primeros Produktions — One of the largest festival catering operations in the Netherlands
- One For The Road — Another major player in Dutch festival catering
- Devilsberg — My own practice, with skin in the game
These aren't passive investors. Primeros and One For The Road are active users of the platform, scheduling their own crews through Crew Craft. They invested because they saw a tool that solved their own operational challenges — and because they believed in the development approach.
What This Demonstrates
Crew Craft is proof that AI-assisted development can produce real, production-grade software used by real businesses. Not a demo. Not a prototype. A platform that has scheduled over 10,000 shifts for some of the Netherlands' biggest festival catering operations.
The key insight: AI doesn't replace human judgement. It amplifies human capability. The navigator still needs to know where to go. But with AI as the driver, you can get there much faster.
If you're interested in what AI-assisted development could do for your project, let's talk.