Comment The collapse of Builder.ai has cast fresh light on AI coding practices, despite the software company blaming its fall from grace on poor historical decision-making.
Backed by Microsoft, Qatar’s sovereign wealth fund, and a host of venture capitalists, Builder.ai rose rapidly to near-unicorn status as the startup’s valuation approached $1 billion. The company’s business model was to leverage AI tools to allow customers to design and create applications, although the Builder.ai team actually built the apps.
Blue-chip investors poured in cash to the tune of more than $500 million. However, all was not well at the startup. The company was previously known as Engineer.ai, and attracted criticism after The Wall Street Journal revealed in 2019 that the startup used human engineers rather than AI for most of its coding work.
Builder.ai grew more forthcoming about the human factor, but the company came unstuck over its finances. It appointed a new CEO, Manpreet Ratia, in February 2025, taking over from founder Sachin Dev Duggal, whom the company credited with “transforming software development through AI-powered innovation.”
It fell to Ratia to inform employees during a May 20 call reported by the Financial Times that the company was filing for bankruptcy as funds abruptly ran out. Builder.ai was reportedly “unable to recover from historic challenges and past decisions that placed significant strain on its financial position.”
While the failure of startups, even one as high profile as Builder.ai, is not uncommon, the company’s reliance on AI tools to speed coding might give some users pause for thought.
The tech industry is generating a tsunami of AI slop, along with a few instances of true value. In the coding world, generative AI can make for useful coding assistants but are often less than helpful when expected to behave like an engineer.
An amusing thread on Reddit titled “My new hobby: watching AI slowly drive Microsoft employees insane” linked to several GitHub threads in the .NET runtime repo in which humans patiently handhold the GitHub Copilot coding agent as it makes mistake after mistake, many of which would make a junior developer blush. It all feels a bit Mechanical Turk, except for coding rather than chess, and the intervention of a human being is all too evident.
One commenter said: “The amount of time they spend replying to a friggin LLM is just crazy… It’s also depressing.”
Last month, Microsoft CEO Satya Nadella boasted that 30 percent of the code in some of the tech giant’s repositories was written by AI. As such, an observer cannot help but suspect some passive aggression is occurring here, where a developer has been told that the agent must be used, and so they are going to jolly well do it. After all, Nadella is not one to shy from layoffs.
The problem highlighted by both the pull requests in the .NET runtime repo and the failure of Builder.ai is that regardless of the wishful thinking from tech giants seeking their next big growth opportunity, startups pitching the latest and greatest innovation, and execs seeking to trim their budgets at the expense of engineers, generative AI tools are not a universal panacea.
Although Builder.ai’s fall has roots in financial mismanagement and forecasts that were arguably over-optimistic, the company was a darling of the generative AI coding industry and an example of how a business can optimize its processes through the application of the technology.
The fact that it wasn’t able to convince enough customers to pay it enough money to stay solvent should give pause to those who see generative AI as a replacement for junior developers. As the experience of the unfortunate Microsoft staffers having to deal with the GitHub Copilot Agent shows, the technology still has some way to go. One day it might surpass a mediocre intern able to work a search engine, but that day is not today. ®