The Rise and Fall of Builder.ai: Unmasking the AI Illusion

Lisa Ernst · 13.02.2026 · Artificial Intelligence · 7 min

In May 2023, Builder.ai looked like the kind of company the AI era was built for: a product that promised to turn ideas into software “as easily as ordering a pizza,” powered by an AI assistant called Natasha. It had marquee backers, a Microsoft partnership, and a story that fit perfectly into the world’s new obsession with automation.

Two years later, that story collapsed into insolvency proceedings, a U.S. bankruptcy filing, allegations of inflated revenues, and investigations that reached former executives. Builder.ai’s downfall didn’t happen because one feature failed. It happened because the gap between what was promised and what could be proven widened—until money, trust, and time ran out.

Quick Summary

What Builder.ai sold: A low-code / “composable” app-building platform that marketed development as fast, modular, and guided by an AI assistant named Natasha.
What critics questioned early: Reports alleged the platform was “human-assisted” far beyond what the branding implied—raising doubts about how much was truly automated.
How it scaled: Big funding rounds (including a $250M Series D in May 2023) and a public partnership with Microsoft amplified credibility.
What cracked first: Revenue restatements, audits, and internal probes—followed by lender action that removed access to cash.
The immediate trigger: A venture-debt lender seized $37M from company accounts; Builder.ai said it could no longer meet payroll and obligations.
The outcome: Insolvency proceedings and a U.S. bankruptcy filing, plus allegations of “round-tripping” revenue and later subpoenas tied to U.S. investigations.
The bigger lesson: Builder.ai became a case study in “AI washing,” governance risk, and why venture debt can turn fast-growing startups fragile overnight.

The Pitch Everyone Wanted To Believe

Builder.ai’s promise was elegantly simple: describe what you want, pick components, and let the platform translate business intent into working software. The company framed this as democratization—software for non-technical founders and small businesses—made practical by automation.

In 2021, Builder.ai launched Natasha publicly as an AI-powered product manager designed to translate ideas into features, timelines, and budgets. The timing was perfect: the low-code movement was already mainstream, and the world was learning that AI could write code in seconds. Builder.ai didn’t just ride that wave—it branded itself as the surfboard.

Act I — Growth, Funding, and “Trust Signals”

Investors don’t just fund products; they fund narratives. Builder.ai’s narrative checked every box: “AI,” “no-code,” “speed,” “cost reduction,” “global scale.” In March 2022, the company announced a $100M Series C led by Insight Partners, with participation including IFC and WndrCo. In May 2023, Builder.ai announced a collaboration with Microsoft—and shortly after, it reported a $250M Series D led by Qatar Investment Authority.

These are what insiders call trust signals: the presence of household names that makes customers and later investors relax their skepticism. If Microsoft is in the room, surely the tech is real. If a sovereign wealth fund is leading a round, surely the numbers are solid. That’s how credibility compounds.

Act II — The “Human-Assisted” Question That Never Went Away

Builder.ai’s core controversy wasn’t that humans were involved. Nearly every serious AI product uses humans somewhere in the loop: QA, edge cases, delivery, customer support, and integration. The controversy was about representation: how much of the product experience was automation, and how much was operational delivery packaged as AI.

Years before the collapse, a Wall Street Journal investigation described Engineer.ai/Builder.ai as “human-assisted” and quoted employees who questioned how the technology was being presented. Later reporting continued to revisit the same theme: the more “magical” the promise sounded, the more customers and journalists wanted to see what was actually under the hood.

The First Cracks: Governance, Probes, and Audits

By 2024, Builder.ai’s public image faced additional pressure from reporting about legal scrutiny around individuals connected to the company. Builder.ai stated those matters related to prior ventures rather than the startup itself, but the reputational risk was real: high-growth companies can survive skepticism—until skepticism reaches the people, the books, and the boardroom.

In early 2025, the company entered a visibly defensive posture: it acknowledged “historic challenges,” began deeper audits, and restated revenue figures. Founder Sachin Dev Duggal stepped down as CEO and was replaced by Manpreet Ratia, who took on a restructuring and credibility repair job at the worst possible time: when creditors, not investors, start calling the shots.

Source: aonhumancapital.co.in

Manpreet Ratia took over as CEO in 2025 during the audit/restructuring phase—when cash, not vision, decides what survives.

Act III — When Venture Debt Meets a Restatement

Equity investors can tolerate uncertainty; debt cannot. Venture debt comes with covenants and triggers, and when financial reporting becomes disputed, lenders may act fast to protect their capital.

In May 2025, reports said a lender seized $37 million from Builder.ai’s accounts after a $50 million debt facility—leaving the company with little usable cash. Around the same period, Builder.ai’s leadership described significant unpaid obligations to major cloud providers, including Amazon Web Services and Microsoft. Once access to cash is restricted, the startup operating model breaks instantly: payroll stalls, vendors cut off service, customers lose confidence, and the story turns from “turnaround” to “triage.”

Builder.ai confirmed it was entering insolvency proceedings, stating it could not recover from “historic challenges and past decisions.” Soon after, the company’s U.S. entity (Engineer.ai Corp.) filed for bankruptcy in Delaware.

The Allegations: Round-Tripping, Inflated Sales, and the “AI Washing” Label

The post-collapse scrutiny went beyond “Was it really AI?” and into “Were the numbers real?” Bloomberg reported that Builder.ai allegedly inflated sales through “round-tripping,” described as circular invoicing with an Indian company, VerSe Innovation. VerSe denied wrongdoing in public statements reported by media. Builder.ai declined detailed comment in several reports while the insolvency process unfolded.

Source: verse.in

VerSe Innovation was named in reporting about alleged “round-tripping.” VerSe publicly denied the allegations.

This is where the story becomes bigger than one company. “AI washing” doesn’t require a fake product. It requires a mismatch between what customers reasonably believe they are buying and what can be demonstrated: automation rates, repeatability, margins, delivery mechanics, and auditability.

What Customers Should Learn (Without Panic)

Builder.ai is an extreme case, but the risk pattern is common: when you buy software as a service from a fast-scaling vendor, you also buy vendor survival risk. If the vendor collapses, you need an exit plan.

Contractual clarity: Who owns the code? Where is it stored? Do you have repo access?
Operational escrow: Can you trigger a handover if the vendor fails (documentation, credentials, dependencies)?
Architecture sanity: Avoid “black box builders” without export paths and maintainability guarantees.
Billing realism: If someone claims “6x faster and 70% cheaper,” ask what part is standardized and what part is custom work.

A Cleaner Timeline (Key Dates)

2016: Founded as Engineer.ai.
Aug 2019: WSJ publishes “human-assisted AI” scrutiny around Engineer.ai/Builder.ai.
Sep 2021: Natasha is launched publicly as an AI product manager (per later reporting).
Mar 2022: $100M Series C announced (Insight Partners; participants include IFC and WndrCo).
May 10, 2023: Builder.ai announces collaboration with Microsoft.
May 23, 2023: $250M Series D announced, led by Qatar Investment Authority.
Early 2025: Revenue restatements, audits; CEO transition from Duggal to Ratia.
May 2025: Lender seizes cash; Builder.ai confirms insolvency proceedings.
Jun 2, 2025: U.S. entity files for bankruptcy in Delaware.
May–Jul 2025: Reporting alleges round-tripping; long-form investigations detail how the company unraveled.
Late 2025: Former CFO subpoenaed as U.S. authorities probe the collapse.

Was Builder.ai “fake AI”?

The most accurate description from reporting is that the product was heavily “human-assisted.” Human involvement is not inherently illegitimate. The controversy centers on how the level of automation and the business performance were represented.

What actually caused the collapse?

The collapse appears to have been triggered by a liquidity shock when lenders seized/restricted access to cash, amplified by disputed numbers, major unpaid obligations, and a loss of confidence that made a turnaround impossible.

What is “round-tripping”?

“Round-tripping” typically refers to circular transactions where companies invoice each other for similar amounts, creating the appearance of revenue without real economic substance. Bloomberg reported allegations of such practices involving Builder.ai and VerSe.

What’s the biggest lesson for the AI industry?

AI claims need measurable proof: automation rates, margins, auditability, and clear boundaries between software and services. When hype replaces verification, governance and finance eventually force a reckoning.