I recently attended a webinar by Elena Verna (Head of Growth at Lovable) on how to land a job at a rocketship startup. 🚀
It was packed with hard truths about networking and “spicy opinions.” But the real spark didn’t happen on the Zoom call. It happened in the follow-up email.
Kasey (from Superinterviews) sent a note to the attendees with a challenge. She shared a “cursed slide,” which is a messy, chaotic visualization of the job hunt, and dared us to “vibe code a better version.”
But I looked at that slide and saw something different. I saw two distinct user personas with conflicting needs...
That’s when the PM brain kicked in. While other people were redesigning the same slide with better colors and cleaner layouts, I was asking a different question: what if the slide itself is the wrong format?
45 minutes later, I had built and deployed two fully functional React apps. Both went live. Both got nominated as finalists (2 of the 8 spots). And I won the grand prize: access to the Superinterviews course worth hundreds of dollars.
But here’s what matters more than the win. I proved something that most PMs still don’t believe: you don’t need engineering approval to validate your ideas anymore. You just need to know how to prompt.
This is a premium post for my paid subscribers. I’m going to give you the exact prompts I used to build both winning apps in what I call “One Shot” builds. No iteration. No debugging sessions. Just clear prompts that generated production-ready code.
If you’ve been waiting for permission to start building, this is it.
The Trigger: A Cursed Slide and a Challenge
The slide was deliberately messy. It showed the job hunt as this chaotic, winding road with dead ends, false starts, and confusion. The visual metaphor was perfect for how job hunting actually feels.
Kasey’s challenge was simple: make it better.
Most people took “better” literally. They cleaned up the design. They made the arrows smoother. They chose better fonts. All valid improvements.
But I looked at that slide and saw something different. I saw two distinct user personas with conflicting needs:
The Operator needs structure. They want a system, a framework, clear steps. They’re the type who buys productivity planners and color-codes their calendar. When The Operator looks at a messy job hunt slide, they think: “Give me a Kanban board. Let me see what stage I’m in.”
The Dreamer needs narrative. They want inspiration, a story, emotional connection. They’re the type who needs to visualize the journey as an adventure. When The Dreamer looks at that same slide, they think: “Show me this as a story I can believe in.”
Traditional product thinking says: pick one persona. Build for them. Ship it.
But why choose when you can build both in under an hour?
The Decision: Build BOTH Solutions
This is where “Prompt-Led” product management gets interesting. I didn’t need to:
Write a PRD and get it approved
Schedule sprint planning
Negotiate with engineering about feasibility
Wait two weeks for a prototype
I just needed to describe what I wanted clearly enough that Lovable could build it.
App 1: The Cat-Reer Mainframe (for The Operator)
A high-tech “operating system” for finding a job. Instead of a board, I built a Cyberpunk Timeline. It shows the career journey as a straight engineering process: It’s about eating, processing, and output. It’s run by 3D cyberpunk cat mascots.
App 2: The Cinematic Storybook (for The Dreamer)
A horizontal scrolling journey that feels like a narrative. I didn’t use a heavy 3D engine. Instead, I built a “Cinematic 2D” experience using pure CSS. As you scroll right, scenes (The Maze, The Gym, The Victory) animate smoothly, turning the job hunt into a visual story.
I built both. I submitted both. And when Kasey reviewed all submissions, both of my apps made the final 8.
Thank you so much for those who supported me on this! 🙌
That’s when I realized something: having two strong entries gave me better odds than having one perfect entry.
The Build: 45 Minutes From Concept to Deployment
Let me walk you through how this actually happened.
I opened two Lovable projects. I wrote two prompts. I hit generate. Both apps built successfully on the first try.
No debugging. No “try again.” No emergency fixes at 2am.
This is the part that sounds too good to be true. But it’s exactly what happened. And the reason it worked is because I’ve spent the last year learning how to write prompts that give AI tools exactly what they need.
Most people write prompts like they’re texting a friend: “Make a job hunt tracker that looks cool”
I write prompts like they’re technical specifications: “Create a Kanban board with glassmorphism cards, salmon accent color (#FA8072), populated with specific data, animated with hover states that reveal additional content.”
The difference? The second prompt gives Lovable zero room for interpretation.
The “Double Tap” Strategy: How I Hacked The Vote
After submitting both apps, I sat back and waited for the judging.
When the finalists were announced, I saw my name twice. Both apps made the cut.
That’s when the strategy shifted. I wasn’t just competing anymore. I was running a campaign.
Here’s what most people don’t understand about competitions: A PM doesn’t just ship. A PM ensures distribution.
I messaged everyone in my network with a specific ask: “Hey, I’m a finalist in the Lovable challenge. I actually have TWO apps in the final 8. Can you vote for Option 1 AND Option 2?”
This wasn’t an accident. This was deliberate product strategy.
By framing my entries as a “Product Suite” (System + Story), I captured 25% of the available ballot spots. While other finalists were splitting votes, I was stacking them.
People loved it. They loved the audacity of building two completely different solutions. They loved that I didn’t pick a lane. They loved that I showed the range.
The result? I won.
Not because either app was objectively the best design. But because I demonstrated something more valuable than good design: the ability to ship multiple solutions quickly and let users choose.
That’s the future of product management.
The Premium Value: The Exact Prompts
This is what you’re paying for. Not the story. The blueprint.