The 200 changes a week come down to one thing: all the context on what to build and why already sits in my head. The AI just executes.
So stop waiting to learn to code. What's the problem you're sitting on that you haven't started building yet?
The hardest part of building software now is knowing what to build. The how got easy.
I ship features in hours and average 200+ code changes a week. I have 0 Python skills.
When you're the domain expert and the builder, that whole cycle collapses. You've lived the problem, so you skip the spec, the sprint planning, the back-and-forth. You just build and iterate.
That's the real solo-founder edge right now. You are the entire feedback loop.
AI runs at the speed of your worst process, your dirtiest data, and your least willing employee.
I spent an afternoon last week in front of PE partners who buy real companies. Most of the economy hasn't moved yet. We're early, and here's the part people keep missing.
So if you're dreaming about what AI does to your margins, budget for the boring work first. I'd bet the winners are the ones who fix the weak links so the agents have something solid to stand on.
Where's the weakest link in your own setup right now: process, data, or people?
And even then, my deal agent Marcus choked in his first month. Hallucinated numbers, missed obvious red flags.
The fix took weeks of wrapping him in a real workflow and pulling the parts that had to be exact into actual code.
Now picture doing that inside a 30-year-old company. The model is the cheap part. The change management, the process redesign, the data cleanup, that's the bill.
@EBogomolovs Iโm just relying on old fashioned dev hygiene. Everything is done in staging databases and staging environments. Nothing goes live unless it passes all the unit, E2E, type, etc tests.
But if you have a method, let me know!
If you havenโt tried โworkflowโ that just launched in Claude desktop, drop what you are doing and test it.
What took me 5 agent chat windows before is now all orchestrated by a single agent.
Iโm using it for multi-hour runs. Just mention workflow in your chat.
@IfeanyiArinze10@pmitu Itโs a very small community so it is spreading through word of mouth. Right now I am at a conference trying to meet more searchers and understand the pain. I havenโt tried large scale marketing yet.
@SMB_Attorney Agreed these are sharp and gritty people. Iโve talked to many trying to buy them. (Iโm a tech bro)
But they also are craving AI insight. I give them my advice for free and have been asked many times to help.
A lot of people will be successful if they prove real value fast.
Scrape hundreds of US business broker sites weekly. I've compiled that data into a broker search tool.
Use it to find the right broker at any intersection of geo and industry.
Also, found some surprising stats in the data:
- 17 firms list half of all US small-business inventory.
- 796 firms (a third of the directory) list only one business per quarter.
- Only 4 firms operate in all 50 states.
Free finder: https://t.co/CD0XQIKuHH
@sanathbhat39 Thank you. The product is Searcher OS - https://t.co/jhdrVUqbkZ
It's built as an end-to-end system for anyone that is trying to purchase a business from $1m to $30m.
There is no MLS for businesses.
Brokers list across dozens of fragmented sites. BizBuySell, Quiet Light, Sunbelt, Empire Flippers, regional broker sites, off-market networks.
To do it right, you are checking 20+ sources manually. Every week.
I did it that way for months. Hours of wasted time.
1/5
@SahilBloom So true. Sometimes intelligence slows you down. An intelligent person considers risk and opportunity costs.
Those things can be paralyzing.