Open loops
Questions
Questions I expect to keep writing around.
These are the questions behind the site. Some will turn into essays. Some will stay open for a long time.
Healthcare organizations
- Which parts of healthcare operations are actually bottlenecked by software?
- Where do organizations need cleaner data before they need better AI?
- How much operational truth is hidden inside reports nobody trusts?
AI systems
- What should an AI tool be allowed to own, and what should it only prepare for review?
- What does reliability mean when the source data is incomplete, stale, or politically loaded?
- How should teams evaluate AI systems that live inside messy workflows instead of clean benchmarks?
Strategy
- Which AI products become workflow infrastructure, and which stay demos?
- Where does build-vs-buy break down because the vendor cannot see the local operation?
- What should investors ask before believing an AI workflow claim?
Writing
- How do you write about real work without exposing people, clients, or details that shouldn’t be public?
- What does a useful case study look like when the client name is not the point?