How a 15‑Person Team Leveraged AI to Earn Tens of Millions Monthly and Let Anyone Build Apps

Project Origin and Open‑Source Growth

In 2023 Anton Osika (CERN physicist) and Fabian Hedin (developer) launched the open‑source project GPT Engineer . The GitHub repository quickly attracted more than 50 000 stars. Programmers worldwide contributed bug fixes, new features and documentation. A 19‑year‑old contributor from India rewrote core modules, resulting in a dramatic performance improvement.

Rebranding to Lovable

To make the tool accessible to non‑technical users, the team renamed the product “Lovable” after a week‑long brainstorming that generated over 50 candidate names.

Release Iterations and User Adoption

Three test releases were performed in 2024. The first caused a server crash, the second produced buggy applications, and the third achieved rapid user acquisition: more than 1 000 registrations within two hours after launch. Reported metrics include 15 employees, approximately 30 000 paying users, about 25 000 new apps created daily, and monthly revenue in the tens of millions of yuan.

Core Functionality

Idea description – the user provides a natural‑language specification, e.g., “an expense‑tracking app with categories, monthly reports and shared ledgers.”

AI‑driven code generation – the system asks clarification questions (theme color, authentication) and then generates a full code base.

Real‑time preview – within minutes a running application with UI is displayed.

Iterative refinement – the user can request UI or logic changes (e.g., “make the save button green and move it to the bottom‑right”) and the system updates the code instantly.

One‑click deployment – a “Deploy” action publishes the app to a public URL.

The generated projects use the React / TypeScript stack that is also employed by major technology companies. The end‑to‑end workflow typically takes 10–30 minutes, compared with weeks or months for conventional development.

Self‑Healing Capability

Lovable includes an automated regression detection and repair mechanism. When a new feature introduces a bug that would normally break existing functionality, the system identifies the conflict and applies a fix, described by users as “a senior programmer monitoring the code 24 hours a day.”

Pricing and Financial Results

Trial – $20 / month

Startup – $50 / month

Enterprise – $100 / month

Custom – negotiated price for large customers

With roughly 30 000 subscribers, annual subscription revenue grew from $4 million to $17 million, while operating costs remained around $2 million.

Funding and Growth Metrics

Seed round in October 2023 raised $7.5 million; Series A in February 2025 raised $15 million, bringing valuation above $100 million. Reported growth figures include $2 million weekly revenue increase and 85 % user retention. Customer acquisition is driven primarily by word‑of‑mouth and social sharing rather than paid advertising.

User Scenarios

Wang Xiaoming, a coffee‑shop owner, built a reservation and loyalty app in one night.

Li Engineer, a software developer, reduced a two‑week backend project to half a day using Lovable.

Zhang, a primary‑school teacher, created a “school‑communication” app that reached 20 000 users.

Li DaYe, a retiree, launched a community‑group‑buying assistant.

Future Outlook

By lowering the coding barrier, Lovable aims to shorten the idea‑to‑product cycle to days, enable solo developers to build complex systems, and expand the pool of non‑technical founders. The authors anticipate exponential growth in software innovation speed over the next five years.