Working note · 2026-10-15
From software to the operations it serves
One path I’m exploring starts with an established software business, moves into AI-enabled workflows, and could eventually extend to owning the operating businesses that use it.
Each step changes the work and the economics:
- Software: Help customers record, coordinate, and manage their work.
- Software with AI: Help perform parts of that work, with outcomes that can be checked.
- Operating-company ownership: Take responsibility for delivering the service itself, potentially using software, AI, and robotics to improve physical operations.
These are possibilities to test, not an inevitable progression. Being good at software does not automatically make a company good at operating its customers’ businesses.
The questions that matter to me:
- Where is the workflow documented well enough to understand what actually happens?
- Which outcomes can be verified, and which depend on human judgment or physical execution?
- Does automation create enough customer value to justify changing an established way of working?
- Is selling a better product the best way to capture that value, or is there a reason to own the operation?
- What new capital requirements, operational responsibilities, and customer conflicts would ownership introduce?
I’m particularly interested in where the transition breaks. A useful AI feature, a dependable automated workflow, and a well-run operating company are three different achievements.
If you’re working through this in a specific industry, I’d enjoy comparing notes.