I shipped a productivity SaaS in 30 days as a solo dev — here's what AI actually changed (and what it didn't)
In 2019, 23.7% of new startups had a solo founder. By mid-2025, that number was 36.3%.
Something structural shifted — and I think I felt it firsthand.
I spent six years building products at companies in Kyiv. I watched features that a single developer could ship in a day get stuck for months in approval chains. The average enterprise PR sits untouched for four days before anyone even looks at it — not because people are lazy, but because process overhead scales faster than teams do.
And yet… products still shipped. Users still came. Revenue went up.
The dysfunction was real — and somehow it didn't matter. That made it more frustrating, not less.
I always wanted to build something of my own.
The blocker wasn't ideas. It wasn't time.
It was design.
I'm a backend-first developer. I can architect systems, write clean TypeScript, ship reliable APIs.
But I can't make things look good.
Hiring a designer for a product with unknown revenue felt like betting money I didn't have on odds I couldn't calculate.
So I waited.
Then the calculus changed.
AI-generated design gave me a starting point — not Dribbble-worthy, but good enough to validate.
AI coding tools handled the parts that usually kill solo projects: boilerplate, tests, repetitive CRUD.
In practice, something that would've taken me ~6 months took about 1 month.
Six months is a bet I couldn't afford.
One month was survivable.
I built Flowly — a workspace for tasks, timers, and analytics.
It's for freelancers who are tired of using 4 different apps just to answer one question:
Where did my week go?
I built it for myself first.
I use it daily.
That's either a great sign — or a selection bias trap. Still figuring that out.
What AI actually changed
Speed
Not across the board — but where it matters. Boilerplate, scaffolding, tests — dramatically faster.
Architecture, data modeling, product decisions — still 100% on me.
Realistically: ~2x–4x depending on the task.
The design blocker
This was the real unlock. Not "AI made me faster" — but "AI removed the reason I hadn't started for 5 years."
The risk threshold
This is the biggest one. A failed 6-month project hurts. A failed 1-month project is survivable.
That changed everything psychologically.
What AI didn't change
Judgment
What to build, what to cut, how to price — still entirely human.
AI executes. It doesn't decide.
Distribution
This is where I'm struggling.
I'm a developer — building feels natural. Distribution feels like guessing.
I catch myself opening VS Code when I should be talking to users.
Shipping code feels like progress. Posting on Reddit feels like gambling.
Not rational — but real.
Where I am now
- Live at flowly.run, with paying users
- 14-day reverse trial (full access, no card → downgrade after)
- Pricing: $8/month annual, $12 monthly
That jump from 23.7% to 36.3% solo founders?
I think it's AI removing the two biggest blockers: time and design.
The window feels real.
I'm trying to use it.
If you've made the builder → distributor shift: what actually changed the game for you?
flowly.run — free tier available, no card required
Top comments (1)
The "(and what it didn't)" is the most valuable half of this post, because that's the honest part everyone skips. My guess at what AI changed: the boilerplate/UI/first-draft velocity went way up. And what it didn't: the thinking - product decisions, the gnarly integration edge cases, debugging your own architecture, and the non-code grind of actually getting users. AI compressed the typing, not the judgment.
The piece I keep seeing eat solo-dev time even with AI is the boring infra layer - auth, billing, deploy - which is exactly the slice I built Moonshift to absorb: a multi-agent pipeline that ships a prompt to a real SaaS on your own GitHub + Vercel with that 20% as defaults, ~$3 flat per build via routing. It doesn't change the judgment part (nothing does), but it gives you those days back. First run's free, no card. Great honest writeup - of your 30 days, what % went to actual product thinking vs plumbing? That ratio is the real story of solo AI building.