5 weeks into building DwellKeep with a full-time job and honestly? those focused 2-hour blocks hit different
way more productive than I expected
curious for those building on the side, what's your sched? early mornings? late nights?
AI prompt cheat code: "Can you help me better understand the tradeoffs of this decision. I'm a beginner."
building an app is 20% coding
80% decisions that sound small but change everything
this week: do I let users add custom tasks
custom = more flexibility, messier data
preset = cleaner experience
leaning toward keeping custom because every home is different
what would you pick?
reddit research day
the number of "how often should i..." posts is wild
also learned that redditors can smell self-promotion from a mile away so... just gonna be helpful for a while #buildinpublic
I'm not a designer, but the @Garmin.com app frustrated me enough to redesign it.
One readiness score. Clear vitals. No clutter.
Would you use this?
#buildinpublic
day 21: dec 23rd - everyone's checking out early and relaxing
me: "just need to fix this one bug..."
Spoiler: it was not one bug
DwellKeep is 90% done which means I'm 50% done (iykyk)
anyone else working while everyone else is on eggnog? 😅 #buildinpublic
Day 20(ish): building my home maintenance app.
When building feels overwhelming, I track micro-wins. This week:
✅ Fixed annoying bug
✅ Got 90% of UX locked
✅ Migrated to improved architecture
Small wins compound. What's your micro-win today? #buildinpublic
What's your favourite place for design inspiration? Currently checking out Mobbin but would like to know what else is out there
day 19: building my home maintenance app
there is one book I keep going back to - "The Dip" by Seth Godin.
You don't even have to read it, just remember there is always a phase where where progress slows and commitment is tested. Focus on getting through.
#buildinpublic
did it! always helpful to have someone watching for that extra push
day 18: building my home maintenance app
Took me a while but realized the value of #buildinpublic
Even with no audience, making a public commitment forces you to deliver. Launched my landing page for dwellkeep.com
Next step: Launch my MVP before year-end
day 17 (?): building HomeTask
have spent too much time improving features.
need to get something out the door ASAP.
tomorrow plan to launch my landing page.
#buildinpublic
5. Interrupt early. If it's going the wrong direction, stop it and redirect
The difference between vague and specific prompts is night and day.
#buildinpublic
3. Watch your examples. AI pays close attention to every detail you show it
4. Give it images. Screenshots and mockups make a huge difference for UI work
👇
How you prompt AI is crucial. While building my app these are the most important learnings:
1. Be super specific. Don't say "add tests" - say "write a test for the logout edge case without mocks"
2. Explain why it matters. Context helps the AI understand your goals better
👇
day 16: building HomeTask
today's question: build notification system from scratch or use a service?
Chose: Roll my own with local scheduling.
Why?
→ No API costs
→ Works offline
→ Full control over timing
Probably overengineering. We'll see. 🤷♂️
#buildinpublic
day 15: building HomeTask (home maintenance app)
Planned to build notifications today.
Ended up going down a rabbit hole of architecture changes.
#buildinpublic
does anyone else notice AI always misspells "memorize"?
Launching a proper beta and getting real user feedback. I keep polishing features instead of putting it in front of people.
If someone feels it's invasive they can always ignore.
Shows you're paying attention and care about the experience. I'd keep it short though, maybe offer a discount or free month as a 'sorry for the hassle' gesture.
day 14: building HomeTask (home maintenance app)
Made some progress evolving the UI. Still further refinement needed. (left: before / right: after)
Next steps: building out notifications
#buildinpublic
day 13: building my first app to solve a real problem - keeping track of your home maintenance.
dialing in the workflow: Claude for features + Cursor to write code. Surprisingly good combo.
next experiment: testing if the Claude Chrome plugin can handle my QA. We'll see.
#buildinpublic
Congrats! Great to see the progress
day 12: building my first app to solve a real problem - keeping track of your home maintenance.
Day in the life. Early morning wakeup, dog isn't happy
#buildinpublic
day 11: building my first app to solve a real problem - keeping track of your home maintenance.
All core features built. Time to cleanup and launch a beta. Nice to see the progress but I'm sure this will take longer than expected.
#buildinpublic
Ended up being a simple copy/paste error in my code. They're always simple issues.
day 10: building my first app to solve a real problem - keeping track of your home maintenance.
Some days you create more bugs than you solve. this is one of those days.
Close the laptop and come back tomorrow.
#buildinpublic
That's right, those are placeholders for images / icons. Appreciate the feedback!
day 9: building my first app to solve a real problem - keeping track of your home maintenance.
what layout would you rather use to track your home maintenance tasks? task list or calendar?
#buildinpublic
My plan is to wrap up my onboarding flow and then get a beta out the door to get user feedback. Definitely would be helpful early on!
