Nobody tells you, when you're deep in a codebase optimizing a training loop, that the hardest technical problem you'll ever solve is convincing someone to pay you before the product is finished. I learned that the hard way building Soletechnix.
The skills that transferred, and the ones that didn't
Being a strong engineer gave me a real head start on some things:
- Scoping accurately. I could tell early whether a client's ask was a two-week build or a two-month rabbit hole, and price accordingly.
- Talking to technical buyers. When a prospect's CTO asked hard architecture questions, I could answer them directly instead of routing to "someone on the team."
- Shipping fast, badly, on purpose. Engineers are trained to avoid technical debt. Founders have to actively choose it, constantly, to survive.
What didn't transfer at all: sales, cash flow discipline, and — the one that humbled me most — knowing when not to build something. Every instinct I had said "I can build that." The market doesn't care what I can build. It cares what people will pay for, repeatedly, before the money runs out.
Three decisions that mattered more than any line of code
- Getting DPIIT recognition early. It sounds like paperwork. It changed how every serious conversation with investors and enterprise clients started — credibility you don't have to argue for.
- Saying no to the first "big" client whose ask didn't fit anything we were good at. Taking that contract would have made year one's revenue look great and year two's roadmap look like someone else's company.
- Shipping Basium AI, Facultify, and NPCtrips as real, live products instead of case studies. A pitch deck full of concepts convinces nobody. Three URLs that work convince almost everyone.
Technical ability gets you in the room. It has almost nothing to do with whether you stay in business.
What I'd tell an engineer thinking about doing this
Build the smallest possible version of the thing that someone will pay for, and go find out if they will — before you write the architecture doc you're excited about. The architecture doc can wait. The revenue conversation can't.
I still write code most days. But the job changed underneath me the day I became responsible for whether there'd be a next month to write it in.