“You need to make software that works for a business owner who spends 95% of their time in the field, has a family, and isn’t trained in accounting, in an industry where you need to understand your balance sheet from day one. In most industries you don’t have to do that until you get much larger. It’s gnarly, but that’s what makes it so rewarding.”

“Working for the real economy felt more tangible to me than the abstract problems being solved by AI today. Small businesses make up so much of our economy, and working with them shows you what technology actually needs to be built. The chance to be on the ground in these messy environments, literally and physically, is deeply enticing, and I think unique.”

“I’m building a self-healing loop that improves our code base so agents run faster, and engineers can ship to customers quicker. It’s greenfield. I haven’t found a company that’s mastered improving a code base with an agent, and no one’s published it, so I’m building a lot of it in house. I’m used to well-scoped work, so it’s exciting to flex the muscle of independent research, experimentation, and actual implementation.”

“We have an incredibly broad product serving so many use cases. The technically challenging part is the sheer scope, and finding a way to build and maintain all of it. There’s a lot of marketing babble in how companies talk about AI tooling, so we’re taking a really grounded, empirical approach to figuring out what actually moves people through product faster.”