Research
Research notes.
Experiments, findings and use cases from the lab, published openly. Many of them become entries in the library.
Latest
How We Build Persistent Context Across AI Systems
A practical way to carry your goals, constraints and decisions from one AI session to the next, without re-explaining your business every time.
Turning Repetitive Business Operations Into Agent Workflows
A way to decide which recurring tasks in a small services firm are worth automating and which are worth deleting.
The Difference Between Using AI and Building AI Infrastructure
Most teams use AI as a tool. Some have quietly built something that needs owners, tests and version control. How to tell which one you are.
Why Most Prompt Libraries Become Useless
How to tell why prompt collections decay, and what the ones that last have in common.
An Opportunity Map for AI Services to Local Businesses
Where AI services for local businesses may fit, which problems buyers can already see, and which only look attractive from outside.
Evaluating Workflow Reliability Before You Trust It
A lightweight way to test multi-step AI workflows: a handful of known cases, drift checks and a list of the ways they usually fail.