M.S. Engineering Β· Exercise Science PhD (ABD).
Personal trainer by day, indie hacker by night ποΈπ»
6 products built, 1 with paying users. From Korea π°π·
How a personal trainer ended up indie hacking:
It started with one client, mid-session.
He's a developer at Krafton, the company behind PUBG. I mentioned I'd dabbled in coding back in grad school but never kept it up.
He said: "You should try vibe coding. It's gotten insanely good."
I was skeptical. A few years ago I tried AI-written code, and it was a mess. Errors everywhere, nothing actually ran.
But I gave it a shot.
He was right. It was a completely different world.
So I started building between PT sessions and in every spare minute.
Fast forward: 6 apps shipped.
Next up: the first app I built, and why it failed. Follow along.
Thank you, that means a lot! Just checked out https://t.co/3ncAJzh65p and it feels like everything a small team or founder needs, packed into one place. Super interesting!
And you're a professor! In Korea, professors are basically gods to grad students lol. As a former grad student myself, huge respect.
I'm a personal trainer by day, indie hacker by night.
Anyone else building on the side while working a completely different job?
Tell me your day job and what you're building π I'll check out every single one. Let's connect!
@addy_devx Let's connect! Just checked out https://t.co/NboCRqJtak and matching WASSCE scores to programs, scholarships, and mentors all in one place is such a socially valuable and rewarding problem to solve.
@pablonasarre__ Β‘Hola! Let's connect! Just checked out https://t.co/e3xSiAnpIZ and the "see your site first, pay only if you like it" model is so cool. You can really feel the confidence in your work. Respect!
@Atulo1o Great to meet you! Just checked out https://t.co/95w88NUHxv and as a working personal trainer, I found it really interesting. Have you ever worked as a trainer yourself, too?
@MorozVladyslav Let's connect! Just checked out your website, https://t.co/TxCnztzjfs, and wow, for a backend dev, your UI design is super cool. Did you design it yourself?
Thank you! I'm not an engineer anymore, though.
I got my M.S. in Materials Science and Engineering, then went on to an Exercise Science PhD (ABD), and now I work as a personal trainer.
But training has a built-in limit: my income only grows with my hours. So I wanted to build something that makes money even while I'm coaching.
@alokdataeng Hi Alok! Oh really? I've heard great things about Cursor. I'll give it a try, thanks! And a web-based game sounds fun. What kind of game is it?
One of the hardest parts of app dev for me: responsive design.
Making it work on every phone, foldable, and tablet is brutal. I'll get a screen looking great on one size, and then it completely breaks on a different aspect ratio.
How do you all handle this? Any tips or tools that actually help?
@alokdataeng He's one of my personal training clients, so we were in the middle of a training session when he said it. Looking back, that one conversation changed a lot for me.
How a personal trainer ended up indie hacking:
It started with one client, mid-session.
He's a developer at Krafton, the company behind PUBG. I mentioned I'd dabbled in coding back in grad school but never kept it up.
He said: "You should try vibe coding. It's gotten insanely good."
I was skeptical. A few years ago I tried AI-written code, and it was a mess. Errors everywhere, nothing actually ran.
But I gave it a shot.
He was right. It was a completely different world.
So I started building between PT sessions and in every spare minute.
Fast forward: 6 apps shipped.
Next up: the first app I built, and why it failed. Follow along.
I've used RAG once, in an app I built (shelved now). Snap a photo of a storefront sign, and it estimates that type of business's average revenue and operating profit. Snap a regular product, and it estimates the margin.
It all started from curiosity: how much would I make running that shop? How much do they keep from selling one of these?
@michaeldeayaw Totally agree, the constraint keeps me focused. Right now it's a web project launching this week, plus an app I'm aiming to ship by the end of the year.
Hours-wise, about 30hr a week, mostly late nights and the gaps between PT sessions. What about you? What are you building?
Honestly, just one: "Review this critically."
I knew LLMs lean toward telling you what you want to hear, but I didn't realize how much. Critical thinking was the most valuable thing I learned in grad school, so I tried asking AI to critically review my own ideas.
It opened up a whole new world. Way more objective, way more accurate feedback. My product quality jumped from that point on.
Here's a habit of mine. Whenever I walk past a shop, I wonder: what's their MRR? When I see a product: how much margin do they make on each one?
Getting the picture?
Yep. Just like Cal AI does calories, I built an app where you snap a photo of a storefront sign or a product. It estimates the cost like zerob does, or uses the industry's average operating margin, and estimates the shop's monthly revenue.
Working name: BizScan (Video below: its onboarding flow)
Under the hood: OCR reads the sign or product, then an LLM pulls from Korea's national small business database using RAG.
But I shelved this one too. Why? More next time.
How a personal trainer ended up indie hacking Β· Ep. 5
Quick recap:
Cal AI: snap a photo of your meal, get the calories.
zerob: buys the ingredients of store-bought food and calculates its real cost.
So what idea did these give me?π
From my own experience (6 products built with AI, 1 paying): yes, up to a point.
Getting to an MVP without a real coder is very doable now, as long as you have solid product sense.
But once real users show up, it's worth bringing in an expert at least a few times for security, performance, and UI/UX. That's the part that keeps a product alive as it grows.
So you can start without one. Just don't plan to scale without one.
@SuiBuilds Same. What helped me: run /insights to see where your tokens actually go, then keep the default effort low and only drop "ultrathink" on the turns that really need it. Lasts way longer that way.