Pinned Tweet
It started as a little side project for a teacher friend of mine. Just wanted to make her job easier and save her some time.
She tried it and loved it. I got a little too excited and ended up putting it on Google Play. LOL 😂
It's called FormationBoard — a simple tool for planning stage formations, previewing transitions with music, n exporting everything as images or videos.
play.google.com/store/apps/d…
#BuildInPublic #IndieDev
2
2
68
Introducing AgentPilot Pro™
> runs your whole business 24/7
> fully autonomous*
> zero human effort**
*asks for approval at every step
**you review every output, fix the edge cases, and explain to customers what happened
Now with 40% more agents.
1
37
A 145B open-weight model that runs on a laptop. That's what Edge0 shipped last week: ~10B active params, ~29 GB of memory.
If local models keep getting this good, what's the first thing you'd stop paying an API for?
1
1
29
Perplexity tip: before you build your next app, ask:
"Find 20 recent Reddit posts and App Store reviews complaining about [app category]. Group the complaints. Which one has no good fix yet?"
That's your feature list. Or your next app.
34
Simon Willison wrote up the "external sandbox" pattern for coding agents: isolated box, only the folders it needs, network through a proxy.
Not ready for the full setup? Start with one line:
"Only touch files in ./app. Ask me before any install, delete, or network call."
31
Founder tip: stop selling "AI automation" to everyone.
> nobody wakes up wanting automation
> generic offers get compared on price
> you end up doing custom work for free
Better move: pick one boring job in one niche (like invoice chasing for clinics) and sell the outcome.
1
28
5,845 of 10,646 startups on TrustMRR made $0 in the last 30 days.
Before your next AI side project, paste this:
"Here's my idea: [idea]. Name 3 kinds of buyers who'd pay this week, where they hang out, and the one message that gets a yes. If you can't, tell me to kill it."
1
1
64
unpopular opinion:
most indie hackers don't have a marketing problem
they have a nobody-wants-this-product problem, and "learning marketing" is just the most comfortable way to avoid admitting it
marketing amplifies demand, it doesn't create it
if 100 of your target users saw it and nobody paid, that wasn't a traffic problem
that was your answer
1
36
Everyone is selling "AI side hustle" courses.
The people actually making money aren't teaching. They're quietly solving one boring problem for one niche, with AI doing 80% of the work.
Boring.
1
2
72
What's the fuss about paying for AI wrappers?
The open-source versions are often a GitHub search away. Whisper for transcription. Ollama for local models. n8n for automation.
What's one paid AI tool you'd never replace with open source?
1
1
69
n8n tip: stop checking your app reviews by hand.
Build one flow:
New App Store review → AI tags it (bug / feature / praise) → bugs go to your issue tracker, praise goes to a "testimonials" doc.
20 minutes to set up. Runs forever.
90
Add these 3 things to your App Store page before spending a cent on marketing:
1. A first screenshot that shows the result, not the UI
2. Your #1 keyword in the subtitle
3. A reply to every 1-star review
Free. Most indie apps skip all 3.
2
1
70
Introducing LaunchDay.
Post your app on Product Hunt, get 400 upvotes from other founders, and 0 users.
Now with AI-generated comments from people who will never open it.
Free. Like your launch.
3
4
67
Most indie apps don't die from bad code. They die at onboarding.
Paste this into your AI coding tool:
"Act as a first-time user. Walk through onboarding step by step. Flag every screen where you'd quit. Suggest the smallest fix for each."
1
58













