Business Student Ships SaaS in 10 Days Using Claude Code

Ten days ago, the concept of a webhook was foreign. Last night, quidkit — a Next.js + Supabase + Stripe starter kit complete with authentication, subscription billing, and documentation — was published. This morning, its creator is on holiday. The architect? A business management student with no deep coding background.

This narrative challenges the conventional wisdom that building a functional SaaS product requires years of dedicated software engineering experience. Instead, it highlights a new paradigm where individuals with strong product vision and management skills can leverage advanced AI coding assistants to bridge the technical gap. The student, who admits they "can't really 'code' in the way that word usually means," managed a "very fast, very literal developer that lives in my terminal." This capability fundamentally shifted what was buildable for them.

This is an honest account of what was built, the workflow with the AI, every bug that threatened progress, and the associated costs. It offers a granular look at how AI is democratizing software development for non-traditional coders.

The Product: Quidkit - A SaaS Starter Kit

Quidkit is not just a simple script or a proof-of-concept. It is a fully functional starter kit designed to accelerate the launch of new SaaS products. The core technology stack includes:

• Next.js: A popular React framework for building server-rendered and static web applications, providing a robust foundation for the frontend.
• Supabase: An open-source Firebase alternative, offering a PostgreSQL database, authentication, storage, and real-time subscriptions.
• Stripe: The industry standard for handling online payments and subscription billing, integrated seamlessly into the product.

The product includes essential features that often consume significant development time for early-stage startups:

• Working Authentication: Secure user sign-up, login, and password management, powered by Supabase Auth.
• Subscription Billing: A complete Stripe integration enabling recurring payments, managing customer subscriptions, and handling payment failures.
• Documentation: Comprehensive guides and explanations for users and developers to understand and extend the starter kit.

The ambition was to create a product that could be shipped and used, not just a theoretical exercise. The student aimed to replicate the core components of many successful SaaS businesses, enabling new founders to bypass the initial, often steepest, part of the development curve.

The AI Workflow: Managing a Literal Developer

The development process was fundamentally different from traditional coding. The student acted as a product manager and a prompt engineer, directing Claude Code, an AI coding assistant, to generate the necessary code. This involved a highly iterative approach:

• Prompting for Features: Each feature or component was requested through detailed prompts. For example, a prompt might specify the need for a user profile page that fetches data from Supabase and allows updates.
• Code Generation and Integration: Claude Code would generate the relevant code snippets for Next.js components, API routes, or Supabase functions. The student then copied and pasted this code into their project structure.
• Debugging and Refinement: This was the most challenging and time-consuming part. When code didn't work as expected, the student had to articulate the error message or the incorrect behavior back to Claude Code, often requiring multiple iterations of prompts and code regeneration. This is where the analogy of managing a "very literal developer" becomes apt – the AI executes instructions precisely, but understanding its limitations and guiding it through complex logic requires significant human oversight.
• Webhooks and Backend Logic: Understanding concepts like webhooks, which are crucial for integrating services like Stripe for payment notifications, was learned on the fly through AI assistance. The student would ask Claude Code to explain the concept and then generate the necessary code to handle these events.

The student’s business management background proved invaluable. They approached development with a structured mindset, breaking down the large goal of a SaaS product into smaller, manageable tasks that could be delegated to the AI. This is akin to managing a junior developer or an offshore team: clear requirements, consistent feedback, and rigorous testing are paramount.