> ## Content Index
> Fetch the complete content index at: https://www.therecursive.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# Lovable’s Vladimir Arustamian on Why AI Won’t Make Startup Success Any Easier
- URL: https://www.therecursive.com/lovables-vladimir-arustamian-on-why-ai-wont-make-startup-success-any-easier/
- Published: 2026-11-17T10:50:33.000Z
- Updated: 2026-11-17T11:01:01.000Z
- Description: As AI makes software development accessible to people without technical background. Vladimir Arustamian of Lovable discusses how founders can turn ideas into products and why distribution remains a major challenge.
- Author: Teodora Atanasova
- Tags: Lovable, vibe-coding, Crossroads, Interview

Launching a software company no longer necessarily requires an engineering team and substantial upfront funding. [AI-powered development platforms](https://www.therecursive.com/tag/vibe-coding/) are already changing **who gets to build software**. But this transition also raises questions about what it takes to build a successful business. Some argue that starting a company has never been easier, while others warn that the market could become saturated with products that nobody actually needs.

**Vladimir Arustamian**, Tech lead at [Lovable](https://lovable.dev/?ref=therecursive.com), shares his perspective on the opportunities and challenges of this shift, and what entrepreneurs should keep in mind when turning an idea into a business, following his appearance at [Crossroads](https://www.startupbulgaria.eu/crossroads?ref=therecursive.com) in Sofia earlier this year.

For him, joining the company meant moving from building his own startup to working at what he considers the frontier of AI. *"Before Lovable I was building my own startup. It was growing, but linearly, and I wanted to work at the frontier of AI,"* he tells *The Recursive*.

> "To me, Lovable is the most AI-pilled company in Europe, and one that genuinely innovates in AI rather than just applying it. I reached out to a few folks at Lovable, saw that my vision was fully aligned with the company, and joined."

## From business idea to working software

Founded in Stockholm in 2023 by Anton Osika and Fabian Hedin, **Lovable emerged from the idea that building software should not be limited to people who know how to code.** The company grew out of GPT Engineer, an open-source project created by Osika that explored how AI could turn natural-language instructions into working code. Building on that foundation, the founders launched Lovable as a platform enabling users to create websites and applications simply by describing what they want to build. 

"*If you have an idea but don’t know where to start, brainstorm it with Lovable first*", says Arustamian. For him, one of the most interesting developments is how easily users can incorporate AI into their own products. One example Arustamian highlights is Retro Knights Arcade, a business created by a **founder who identified an empty spot in the market** and decided to turn it into an arcade.

> "A founder noticed an empty space, borrowed $50K worth of arcade machines and gave himself three months to make it work. With no developers, he used Lovable to build everything the business runs on: bookings, memberships, staff training and live leaderboards. Retro Knights Arcade now brings in $30K a month."

Another **example comes from the nonprofit sector**. *"The founder of WeGlow has never written a line of code, yet he built its donation platform in a day. More than 100 nonprofits now use it and have raised over €1M."*

The technology is not limited to early-stage businesses. **Established companies are also experimenting** with building their own software instead of purchasing existing solutions. *"Inside companies, Bulgarian company Payhawk built a performance-management system that’s replacing their HR system, along with an OKR tool that replaced another product. Atonom replaced a $40,000-a-year Salesforce contract with a CRM built in Lovable."*

## Starting a business has never been easier. Succeeding is another story.

While AI is reducing the technical barriers to entrepreneurship, Arustamian is careful not to be too optimistic that building a successful company has become equally easy. One of the biggest mistakes entrepreneurs make, he argues, is **confusing positive feedback with genuine customer demand.**

> "When you ask people whether they like your idea or would pay for it, most will say yes. That doesn’t mean they will actually pay."

Even with faster development and lower initial costs, founders still need to follow the **regular steps of establishing a business**. *"Founders also underestimate how hard distribution and funding are. It’s usually far harder than you expect, so you need some kind of unfair advantage,"* he explains.

> "Starting a business has never been easier, but building a successful startup is as hard as ever."

This raises an important question for the next generation of entrepreneurs: if software becomes easier to create, what will differentiate one startup from another? According to Arustamian, **the traditional sources of competitive advantage remain relevant.**

> "The usual moats still apply: unique data, domain expertise and distribution. People also underestimate form factor and UX."

Arustamian advises founders to start with a demo, test it with at least ten potential customers, gather feedback to better understand their needs, and refine the product before assessing their willingness to pay. He points to personal AI agents as an example of how a new way of delivering technology can create value even when the underlying capabilities already exist.

Lovable has attracted attention for the speed at which its team develops and executes ideas. Arustamian attributes this to three factors: **culture, talent, and the pace of information-sharing.** *"In my experience, it matters a lot to spend time together, in person. At Lovable, people genuinely like working with each other, and keeping that as you grow takes real effort."*

> "A good startup should feel almost like a cult."

## When AI builds the product, who is responsible for security?

Making software development more accessible also raises **questions about reliability**, particularly when applications are built by people without traditional engineering experience. Arustamian emphasizes that security becomes essential as soon as a product reaches real users.

> "The moment you put software in front of real users, security is the product: it has to work, protect people’s data, and be dependable."

He explains that Lovable provides built-in security capabilities, but founders still need to take responsibility for **how their applications are configured and tested**. *"Your platform should handle most of that for you. Lovable builds in authentication, access controls, encrypted secrets, and continuous security scanning, and checks your app before every publish."*

However, he cautions that not every application is automatically ready for customers. His recommendations include **testing applications with two separate accounts to ensure users cannot access each other's data**, keeping API keys out of the frontend, and resolving critical issues before publishing.

Arustamian also encourages founders to test how their applications behave with unexpected inputs and on mobile devices, enable project monitoring, and **use the platform's Trust Center to communicate their security practices.**

> "Security matters more the bigger your company gets."

Ultimately, Arustamian’s message is that while AI has put the tools to build a software company into virtually anyone’s hands, **the fundamental rules of entrepreneurship still apply**. Understanding customers, finding the right market, building secure and reliable products, and creating something people are willing to pay for remain as important as ever.