- Like Magic: AI newsletter
- Posts
- Influencer Marketing, But Make It AI!!!
Influencer Marketing, But Make It AI!!!
AI is here to stay!
Hi fam!
This week, we spotlight the tools redefining content creation, from Napkin AI’s image wizardry to Lovable’s app-building sorcery. Plus, meet Hypefy, the Balkan-born AI startup that’s turning influencer chaos into campaign clarity. Whether you're a creator, marketer, or just AI-curious, this issue is packed with stories that prove: the future isn’t coming — it’s already here.

“Don’t Negotiate with Influencers” — The Wild Story of Hypefy and the Rise of Steve
"You don’t need to negotiate. Just pick your people, pay the price, and run the campaign. Done. That’s Hypefy." – Stjepan Zelić, CEO
Imagine you’re running a brand. You want some influencers. You don’t want drama. You don’t want to DM 25 creators just to get ghosted by 23 of them and overcharged by the other 2.
You want it fast, fair, and ideally... AI-powered.
Well, now you’ve got Hypefy.

Meet Steve: A Guy from Promina with a Brain Wired for AI
Stjepan Zelić — or “Steve” if you’re lazy like me — is not your typical startup founder. He’s a math guy. An electrical engineer. Born in the hills of Promina, raised on stories of resilience, and now leading a high-tech AI startup that's fixing the mess that is influencer marketing.
He didn’t come from Silicon Valley. He came from a village called Zelići, with a detour through Ireland, Pašman, and an apartment in Zagreb. He built Hypefy during COVID lockdowns while others were busy learning to bake sourdough or bingeing Tiger King.
"Hypefy was conceived during the COVID lockdowns, a period when everyone was searching for something meaningful to do. It turns out, some of the best ideas are born in times like these.“
Translation: When the world hit pause, Steve hit play.
The AI Weapon That Finds Influencers So You Don’t Have To
Hypefy was founded by a trio of childhood friends and engineers — Stjepan Zelić (CEO), Filip Špiranec (CTO), and Davor Šibenik — who turned their pandemic brainstorm into one of the hottest AI startups in the Balkans.
Hypefy is simple on the outside. But under the hood, it’s all jets and lasers.
Here’s the pitch:
Tell Hypefy your ideal customer
It scans every influencer in the world (yes, literally)
It finds the top 50 creators with matching audience and vibes
It shoots off automated invites
And boom — campaign launched, no negotiations needed
"You don’t negotiate. You pick. We match. We run."
And the coolest part? They even scan the influencer’s actual content.
Using Azure Vision and a bit of AI sauce, Hypefy analyzes the last 10–15 posts of each creator. It filters for things like brand safety, swearing, content themes — you name it.
So when a brand says:
"We want family-friendly creators who don’t swear and post about food…"
Hypefy delivers just that — at scale.
The Playboy Story: From Facepalm to Deal Closed
One of Steve’s funniest stories? He sent a sales guy to pitch Playboy...
The guy didn’t realize it was that Playboy.
He thought it was some small brand named “Playboy-something”.
No prep. No slides. Just vibes.
"He walks into the meeting and says, “I’m from Hypefy,” and they reply, “We’re Playboy.’”
But guess what? They still landed the account.
Because even global brands are tired of the influencer chaos. Hypefy made it easy. One platform, clean data, fast results.
They now help Playboy choose their next Playmate via influencer analytics. Wild.
Real Talk: What Kind of Results Are We Talking?
Steve broke it down like a Balkan Gary Vee:
"If you used to get 5,000 views with influencer marketing, now you’ll get 10,000. If you got 100,000, now you’ll get 300,000. At least double."
Brands like Nivea and Jaffa have seen double or triple performance after switching to Hypefy. Not because Hypefy has some voodoo. But because it matches the right content to the right creator at the right price, and does it at scale.
They now manage campaigns for over 250 brands, with 70–80 active at any time.
Why It Works: AI + Automation + No BS
Let’s recap how Hypefy is different:
No negotiations. Fixed prices. Period.
Recommender engine. AI suggests creators based on actual fit, not follower count.
Visual analysis. It actually watches and understands videos before suggesting a match.
Scaling power. You don’t hire 3 influencers, you hire 100 — automatically.
Smart filtering. Age, location, niche, personality traits — all considered.
"Offer a fixed price and scale it from 5 to 100 influencers. The performance will explode."
From Zagreb to the World (and Maybe Your Brand?)
In 2025, Hypefy raised €1.75 million from Interactive Venture Partners, the same people managing billions for Thomas Peterffy, one of the richest men in the world.
They’re now planning to take Hypefy global — from Slovenia to the U.S., one automated campaign at a time.
"First, make it look like you’re a legit company. Then you actually become one."
Hypefy Founding Team
Final Takeaway: Don’t Sell the Dream. Automate It.
Hypefy didn’t reinvent marketing.
They just made it not suck.
They didn’t promise 100x ROI with chatbots.
They just said — here's your audience, here are the creators who influence them, here’s the price, go.
That’s the beauty of it.
So if you’re a brand, stop wasting time negotiating with DMs.
If you’re an influencer, get matched and paid — without the spam.
And if you’re like me, wondering who’s quietly building the future of advertising from a laptop in Zagreb...
Now you know.
His name is Steve. The startup is Hypefy. And the game just changed.
Find out why 100K+ engineers read The Code twice a week
Staying behind on tech trends can be a career killer.
But let’s face it, no one has hours to spare every week trying to stay updated.
That’s why over 100,000 engineers at companies like Google, Meta, and Apple read The Code twice a week.
Here’s why it works:
No fluff, just signal – Learn the most important tech news delivered in just two short emails.
Supercharge your skills – Get access to top research papers and resources that give you an edge in the industry.
See the future first – Discover what’s next before it hits the mainstream, so you can lead, not follow.

I Built an AI Journal Reviewer in 3 Hours — and Lovable Made It Feel Like Magic ✨
Ever had one of those ideas that sound insane at first… then three hours later, you’re staring at a working version thinking, “Wait, did I just build that?”
That was me last night — fueled by espresso, curiosity, and a tool called Lovable.
Here’s what went down 👇
The Idea
As someone knee-deep in science and publishing, I’ve seen how chaotic journal submissions can get. Authors send in manuscripts that don’t follow Author Guidelines, reviewers drown in formatting errors, and editors play email ping-pong for weeks.
So I thought:
“What if AI could pre-review manuscripts before a human ever touches them?”
Basically, a compliance checker for scientific journals. I bet they would be willing to pay for the pre-review service. 💵💵
Each journal would have its own account and uploaded Author Guidelines. When a reviewer receives a new manuscript, they upload it into the app — and AI checks if it follows the rules.
It returns a detailed, four-part report:
General Opinion: accept, revise, or reject.
General Remarks: overall structure and scientific value.
Specific Remarks: grammar, citations, figures — all the nitty-gritty stuff.
Other Problems: ethics, unclear data, missing info, etc.
Basically, a digital reviewer who never sleeps, never complains, and knows the Author Guidelines better than the editors themselves.

The Build
I fired up Lovable, which is like ChatGPT on steroids for app creation.
You describe your idea — and it literally builds the app for you.
I dropped in a detailed prompt describing:
Multi-journal support
Guideline uploads with versioning
Reviewer accounts
PDF uploads
A report generator connected to OpenAI
And Lovable just… started building. UI, backend, database, auth — all scaffolded in minutes.
Then came the magic part — connecting it to the OpenAI API.
A few tweaks, an API key here, a pipeline definition there… and suddenly, I had a working system that could:
✅ Parse Author Guidelines
✅ Analyze manuscripts
✅ Generate structured feedback
From zero to a functioning AI-powered review app — in under 3 hours.
That’s not a typo. Three hours.
The Little Twerks
(Yeah, twerks. Not tweaks. Because this build danced its way into existence.)
Sure, I had to fine-tune some things:
Adjust prompt formatting for the AI model
Handle multi-journal permissions
Format the PDF export
Add the “Copy All Feedback” button

Some corrections were needed as the data upload did not work 😆
But Lovable’s code editor made it ridiculously easy. Every time I asked it to change something, it just… did. Instantly.
No 12-tab debugging marathons. No waiting on developers.
⚡️ The Result
A clean, functional web app where:
Each journal has its own account.
Editors can upload new guidelines as they evolve.
Reviewers upload manuscripts and instantly get an AI-based pre-review.
Reports are exportable as PDFs or copy-paste ready for editorial use.
It’s not just automation — it’s augmentation.
Reviewers save hours. Editors get consistency. Authors get faster feedback.
And me?
I got to build a legit SaaS-grade tool in one evening for about $25 (a minimal one-month subscription).
Verdict
Lovable isn’t just a dev tool. It’s a co-founder that codes.
I give it 5/5 on the LMAI scale —
that’s “Like Magic, AI”.
If you’re sitting on an idea — big or small — stop overthinking it.
Open Lovable. Write your dream prompt. Let the machine handle the rest.
TL;DR:
Built an AI platform for scientific journals to pre-review manuscripts using Lovable.
Zero code → full prototype in 3 hours → connected to OpenAI → works beautifully.
If you love ideas that build themselves, this one’s for you.
Choose the Right AI Tools
With thousands of AI tools available, how do you know which ones are worth your money? Subscribe to Mindstream and get our expert guide comparing 40+ popular AI tools. Discover which free options rival paid versions and when upgrading is essential. Stop overspending on tools you don't need and find the perfect AI stack for your workflow.

Messy Ideas, Clean Diagrams — Meet Napkin AI
Ever had a brilliant idea… that just looked boring in plain text? Napkin AI is here to fix that.
This clever little tool transforms your notes, thoughts, or explanations into clean, easy-to-share visuals — charts, diagrams, and infographics — all generated from text alone. Think of it as your pocket designer who never argues about font choices.
What’s Cool About It
Text-to-Visual Magic – Type an idea, concept, or short paragraph — Napkin turns it into a chart or mind map in seconds.
Instant Formatting – It chooses layout, colors, and icons automatically.
Smart Refinement – You can adjust shapes, move elements, or regenerate styles with one click.
Collaborative Sharing – Copy, embed, or export visuals directly for presentations, newsletters, or social media.
Perfect for Creators & Educators – Explainers, storyboards, lesson plans, workflows… you name it.
The Other Side
Limited Styling – You can’t (yet) pick brand colors or fonts.
Simple Outputs – Think “smart sketch,” not “final design.”
Text-Heavy Prompts Confuse It – It prefers concise inputs — short and structured works best.
🧪 Our Test – From a boring workflow to a smart diagram
For our test, we asked Napkin AI to visualize this:
“How AI tools are reshaping education — from personalized learning to automated feedback and creativity boosters.”
Within seconds, it generated a clean mind-map-style infographic with three main branches: Learning, Feedback, and Creativity.
Each branch had neat sub-points and tiny matching icons (books, checkmarks, lightbulbs).
We then added a second test. Generating text and visuals from:
“Workflow of creating a video: idea → script → voiceover → edit → publish.”
Result? A full-blown text with an option to visualize a process that looked presentation-ready.
Sure, it’s not Illustrator-level design, but for something created from an idea under 10 seconds… pretty amazing.

Napkin AI interface turning short prompt into text and clean, minimal infographics.
Final Thoughts
Napkin AI feels like a visual notepad with a brain. It’s fast, intuitive, and produces shareable diagrams that make any explanation feel smarter. The minimalist interface is distraction-free — one text box, one button, instant visuals.
We’re giving it 4/5 on our LMAI scale — because anything that turns messy notes into beautiful ideas deserves applause. It could use more aesthetic input or just more ways to customize.
Now excuse us while we feed it our entire to-do list and see if it can make that look good.

Text generation | Image Generation | LMAI recommends |
|---|---|---|
Like Magic AI NFT 🏞️
Our master plan is to publish an NFT image in each newsletter and hand it out to our subscribers. The earlier you subscribe, the smaller the series are. It's a future collectible, a piece of digital art that captures the essence of this moment in time.
Thank you for being a valued subscriber. Together, let's embrace the magic of AI and creativity!

LMAI149-24102025
Was this email forwarded to you? Sign up here 👇




Reply