Loveable

Loveable emerged as a no-code platform designed to democratize software development. The core insight: if AI can generate code, why do builders need to understand traditional programming? This question attracted Lazar Yovanovich, who had no coding background but possessed the clarity of mind to explore what was actually possible with AI-assisted development. Rather than limiting himself to what he thought was possible, Lazar's lack of technical training became a feature, not a bug—he didn't know that building Chrome extensions or desktop applications "shouldn't" be possible with a web builder, so he just did them.

Lazar's early projects at Loveable validated a radical hypothesis: the bottleneck in AI-assisted development isn't the AI's ability to write code—it's the human's ability to be clear about what they want. In his first week, Lazar could build. By week two, he optimized for speed. By week three, he questioned whether he should have built it at all. This evolution led him to a systematic approach: spend 80% of time planning and 20% executing. Rather than iterating on a single direction, he runs five to six parallel projects simultaneously, trying different approaches, picking the winner, then committing to a detailed plan before final execution. This technique, which emerged from simply having tabs open while waiting for AI to finish, became his core productivity system.

Loveable's growth has been driven by product-led growth—users discovering the tool, building with it, and sharing their creations. Elena Verna, Loveable's Head of Growth, brought Lazar on as the first dedicated vibe coding engineer, recognizing that the fastest way to validate product ideas was to have someone who could ship them at velocity. Internal tools and public-facing products alike (Shopify integration templates, a merch store) showcased what was possible. The community embraced the platform, with people like Whitney, a community manager, discovering undocumented capabilities (like video generation) through experimentation, which later became official features.

The breakthrough was understanding that clarity, not coding skill, was the constraint. Lazar developed a framework for being radically clear with AI: (1) Start with a brain dump using voice dictation, (2) Refine with a typed prompt showing more structure, (3) Find a reference design on Mobbin or Dribbble and attach a screenshot, (4) Find similar code on a template site and attach the actual code snippet. Each iteration adds clarity without adding cost—most tools have free plans. By the time you've built four or five versions, you have extreme clarity on direction and save hundreds of builder credits on refinement.

When things go wrong—code breaks, features don't work—the issue is almost never the AI. It's insufficient context. Lazar combats this by treating the token/context window as a scarce resource. He documents everything: a master plan (10,000-foot overview), design guidelines (CSS-informed visual parameters), user journey (how people navigate), implementation plan (sequence of work), and tasks.md (granular next steps). He then sets agent rules telling the AI to read these files before doing anything. This shifts the burden of context-management from the prompt to documentation. Rather than prompting repeatedly, Lazar simply says "proceed with the next task," because the AI already has the roadmap.

Loveable is the fastest-growing platform in its category, with Lazar shipping both internal and external products. He's proven that a non-technical person can build production-quality applications—Shopify templates, merch stores, feature adoption matrices, custom internal tools with dozens of edge functions. The platform has attracted people across the company who now want their own Lazar. His core insight—optimize for judgment, taste, and clarity, not for raw coding speed—is reshaping how teams think about technical hiring and product building. The future, as Lazar sees it, belongs not to faster code generators but to people with better taste, better judgment, and the clarity to communicate complex ideas to machines. He's optimizing 100% of his time on these skills, confident that the code generation part will only get better.