After some false starts (during which I have not kept up the blogging habit), I finally found something that, per 16 May 2025: Conviction is Everything, I have sufficient conviction in to take a big swing. In time, I will have more to unveil, but here’s a quick recap of the false starts:
- Abortive startup endeavor with high-status collaborators who were willing to offer time and money. Really promising on paper, but the problem is the collaborators were remote, not full-time, and one of whom is really locked in onto his idea.
- Short stint at a crypto startup that’s making a lot of money but is not clearly adding to net social good
- On-again-off-again (booty) calls from a husband-and-wife team trying to build a AI-powered software business to disrupt their own consultancy business
- Worked with an ex-investor/ex-startup founder to teach AI to finance professionals. Learned a lot about communicating to a nontechnical audience and how AI adoption in the enterprise looked like, but no clear “product-shaped” opportunity.
Fortunately, I finally landed on a more conventional startup-shaped idea with a business cofounder and have been busy executing on this. There is a lot to share about this, so here is just some top of mind stuff:
- Have settled into a tech stack that I feel very confident about being able to iterate quickly on. Open-source software + AI coding agents + monorepo — this is really the golden age for programming. I am still in the zero-to-one stage, so the key is to get product in front of actual users and get the feedback loop going
- One clear concern is the market doesn’t seem that big, but we have identified a niche of customers with real pain points and willingness to give us time of day. One upside of all the AI hype is that customers that were previously not receptive to buying software have become much more willing to do so. I think we will close a few contracts real soon and there is opportunity to grow from there.
It’s a good time to be taking more risks and starting one’s own company.