few months ago, my friend get me a freelance client who just wanted to finish his saas product. which was completly vibe coded, it was working but not completed, there had bugs,was full of ai slop and I just fixed and got paid for it, got recommended, get new freelance projects, later making this freelance work as an agency and today we have onboarded a yc backed startup to clean up their code, never thought while started coding that just fixing the products will get us money. A big win for my agency today.
Got hired by a YC startup to clean up their AI slop
byu/InstructionCute5502 inEntrepreneur
Posted by InstructionCute5502
3 Comments
That’s exactly how we stumbled into our niche too. Started fixing one broken MVP for a friend’s fintech startup, then word spread that we could actually ship working code vs the half-baked AI generated mess they had.
The YC connection is huge though – those founders talk to each other constantly. We cleaned up one YC company’s codebase and suddenly had 3 more reaching out within weeks.
Few things that helped us scale this:
– Document everything you fix (we use notion to track all the anti-patterns we find)
– Get testimonials immediately while the pain is fresh
– Focus on specific tech stacks – we only do React/Node shops now
The demand is insane right now. So many teams rushed to ship with AI tools and now they’re paying the price. We’re using Cloudastra Technologies for our cloud infrastructure since we need to spin up test environments quickly for each client cleanup.
Keep pushing on the YC angle – that network effect is real.
What is it like to fix vibe coded MVP’s?
Do you shred the whole thing off and start from scratch?
Is it cheaper for someone to vibecode their idea first and then get it finished/ Production ready?
Would you be able to equate the AI slop to a level of programmer? Like, was it the equivalent to code written by an undergrad or just a decent programmer who was under pressure and cutting corners?
I’m assuming that the YC startup just wanted to get an MVP together to attract investment or validate the idea with initial customers, I imagine that’s going to be the way going forward (and maybe always has been with SaaS start-ups)