Engineer.ai’s AI Was Just Humans the Whole Time

Remember Engineer.ai? Probably not, unless you were following the startup hype cycle back in 2019. This Indian company raised nearly $30 million from SoftBank and others, claiming it had built an AI-powered platform that could whip up 80% of a mobile app in about an hour. The founder, Sachin Dev Duggal—whose official title was apparently “Chief Wizard”—even got on stage at a conference to sell the dream.

Turns out, the magic was just people.

A Wall Street Journal report at the time exposed that Engineer.ai wasn’t using AI to assemble code at all. It was using human engineers in India and elsewhere to build the apps. The company’s chief business officer, Robert Holdheim, actually sued earlier that year, alleging Duggal was telling investors the product was 80% done when he’d barely even started.

The company’s defense was weak. They said they used natural language processing to estimate pricing and timelines, and a “decision tree” to assign tasks. That’s not modern AI. That’s basic software logic. No AI agent was compiling code. No machine learning model was generating features. Just people, doing the work, under a fancy label.

This isn’t an isolated case. A 2019 study by MMC Ventures found that startups claiming an AI component could attract up to 50% more funding than other software companies. And they suspected 40% or more of those companies didn’t actually use any real AI. The .ai domain from Anguilla? Usage doubled in a few years. Slapping “AI” on a pitch deck became a shortcut to investor attention.

The deeper problem is that deploying real AI at scale is hard. It requires massive training data, expensive infrastructure, and top-tier talent—the kind Facebook and Google hoard. Most startups don’t have that. So they fake it until they can maybe build it, or until they get caught.

What’s uncomfortable is how much of what gets called “AI” is actually human labor behind the curtain. Content moderation on platforms like Facebook and YouTube? Some AI, but mostly contractors reviewing violent and harmful content. Customer service chatbots? Often humans typing from scripts. Engineer.ai was just the app-building version of that.

We’ve seen this pattern repeat. The hype cycle rewards the appearance of innovation over actual innovation. Investors get excited, money flows, and the real work—the boring, expensive, human work—gets hidden behind a shiny AI label. Engineer.ai got caught, but plenty of others didn’t. And some probably still haven’t.

The lesson isn’t that AI is useless. It’s that we need to be skeptical when a startup promises to automate something complex overnight. If it sounds too good to be true, it’s probably humans in a room somewhere, doing the work, while someone with a fancy title collects the checks.