Joël Heil Escobar
I'm a founder, product engineer and community builder.
Currently building Attacal : AI agents that attack your app every day and open a pull request for every hole they find, so you never have to think about security again.
a little background
I started a YouTube channel at 10 with Minecraft mod tutorials and tech reviews. At 14 I founded a web dev and video agency that made €20k+ while I was still in high school. Later I studied computer science at CODE University and did ML at Stanford .
Next to uni, I automated the finances of a marketing agency and tried out lots of my own startup ideas. With we built AI agents and a data pipeline that analyzed 1.2M+ TikTok and Instagram videos for Nike and Mastercard. Other ideas became voice agents for car dealerships, a learning app that combines mind maps and flashcards and an AI that plans group vacations.
Most recently, I built and shipped an at SF Consumer Lab in two weeks and made $4k in revenue within four weeks of starting.
YFN
At 16 I ran a podcast asking founders where they failed and met other 16 year olds on LinkedIn. Together we started the Young Founders Network .
In three years I scaled the team from 5 to 100+ volunteers. Today it's Germany's largest community for founders under 25. We have 2,200+ members and have raised €350k+ for entrepreneurial education.
work, startups & projects
2026
2025
2024
2023
earlier
now
AI agents that attack your app every day and open a pull request for every hole they find. AI writes more of our code every month, but pentests still happen once a year. I want every app to have its own attacker.
past
I was the only engineer on a bot-free AI meeting notetaker for Mac and Windows. We went from idea to launch in two weeks. Four weeks in we had 200 paying customers and $4k in revenue, and 800 users after six. Next to the code I ran infra, releases and customer support.
PlayTheHype
I co-founded here: AI market research where Gemini watched 1.2M+ TikTok and Instagram videos and agents turned them into insights for Nike and Mastercard. I also built after watching the team copy comment links into Excel every week. Now a Telegram bot does it.
Stanford
Took machine learning (CS229) and technology entrepreneurship. With three friends I built , a directory of summer students by skills and courses. 60 students joined and teams for one course formed through it. I wrote about the summer.
CODE University
Next to classes I kept starting things. put flashcards inside a mind map, and we stopped it because we never talked to enough users. I built , a map of Berlin's running clubs, on my own. put voice agents on the phones of car dealerships, and the team fell apart when I left for Stanford. started as my capstone and now runs all of Young Founders Network. With a teacher I mentored I built , a TikTok fact checker for kids that Germany's First Lady awarded. My thesis asked whether AI agents can replace marketers.
A marketing agency in Berlin where financial planning ran on memory and copy paste. I automated it and cut errors by about 95%. On the side I managed campaigns for freenet, Gerolsteiner and RheinEnergie that got 3.7M+ views.
At 16 I hosted , a podcast asking founders about their biggest failures. Through it I met other 16 year olds on LinkedIn and we started YFN together. It went from a meetup in an old supermarket hall to Germany's largest community for founders under 25, with 2,200+ members, 100+ volunteers and €350k+ raised. I wrote about leading it.
A startup that replaces internal emails nobody reads with podcasts. I cut audio and video and built the marketing funnel, two and a half years next to school.
Teen projects
I started a at ten after my dad offered a deal: less watching, more making. 150 subscribers, zero fame. At 14 I started building websites for small businesses, which became and made €20k+ before I finished school. In between came an that paid out for five years, a that burned most of my savings and getting my whole school onto . An internship at taught me a new plane takes ten years, and I wanted software.
Currently reading
Read
- The Debt of Time
- Deep Work
- Man's Search for Meaning
- Principles
- The Almanack of Naval Ravikant
- Outperformer
- Why We Sleep
- The Four Agreements
- So Good They Can't Ignore You
- Epiktet / Epictetus - Handbüchlein der Moral / The Enchiridion - The handbook of moral instructions
- Frank Thelen – Die Autobiografie
- The Diary of a CEO
- Factfulness
- Venture Deals
- Burn Book
- Harry Potter and the Cursed Child
- Harry Potter and the Order of the Phoenix (Harry Potter, #5)
- Harry Potter and the Half-Blood Prince (Harry Potter, #6)
- Harry Potter and the Goblet of Fire (Harry Potter, #4)
- Harry Potter and the Deathly Hallows (Harry Potter, #7)
- Harry Potter and the Chamber of Secrets (Harry Potter, #2)
- Harry Potter and the Prisoner of Azkaban (Harry Potter, #3)
- Harry Potter and the Philosopher's Stone (Harry Potter, #1)
- The Minimalist Entrepreneur
- The Hard Thing About Hard Things
- The Subtle Art of Not Giving a F*ck
- Hooked
- Wie man Freunde gewinnt
- Nach ganz oben
- Starting Your Business (Small Business Guides)
- Steve Jobs
- 10xDNA
- Young Money Guide
- React Native
- Elon Musk
Want to read
Did not finish
Life
- be delusionally optimistic, the world works in your favor
- have a bias towards action: if everything seems under control, you're not going fast enough
- if you're the smartest person in the room, switch the room
- we overestimate one year and underestimate five
- rely on systems, not willpower
- ask boldly, revise beliefs quickly
- guard each yes: deciding what not to do matters as much as deciding what to do
- avoid luxury, play the long financial game
- acknowledge luck and risk in achievements, yours and others'
Building products
- sell before you build
- get very close to your customers: work in their office if you can, talk to them daily if not
- build something a few users love, not something many users like
- if it takes more than a sentence to explain, it's too complicated
- work on the idea you think about when you're not trying to think about work
- telling people your idea isn't as dangerous as working on it and never telling anybody
- shipping fast beats the best strategy
Learning
- learn a topic by building with it, not by reading about it
- understand instead of memorizing: active recall and spaced repetition beat rereading
connect
If you're building something interesting, I'd like to hear about it :)