I replaced a $120/year micro-SaaS in 20 minutes with LLM-generated code (blog.pragmaticengineer.com)

In a striking demonstration of the capabilities of large language models (LLMs), a software engineer successfully replaced a $120/year micro-SaaS service, Shoutout.io, with LLM-generated code in just 20 minutes. The engineer sought an alternative after experiencing frustrations with broken billing and lackluster support from the SaaS provider. By leveraging tools like Codex, they developed a personalized solution for displaying testimonials, simplifying the process by utilizing separate JSON files for testimonial storage and deploying it on Netlify. The final product retained the original’s visual appearance while eliminating reliance on a third-party service. This development is significant for the AI/ML community as it underscores the potential of LLMs to streamline software engineering processes, particularly for developers who are comfortable with coding. While the engineer noted that more complex SaaS products would still require significant features beyond basic functionality—such as compliance and real-time analytics—this case illustrates the growing trend of customers opting for custom-built solutions instead of traditional SaaS offerings, especially when those offerings fail to deliver ongoing value. The ease of using AI for such replacements may challenge the profitability and investment strategies of SaaS businesses that do not consistently evolve and maintain high-quality service.