Local AI field notes
The method is public. The client work is private.
Who this is for
Section titled “Who this is for”- Partners and practice managers in small regulated firms.
- IT advisers helping a firm decide what may leave the building.
- Educators teaching AI with student data.
- Anyone running local models on modest hardware.
What is here
Section titled “What is here”- AI readiness questions: five questions for a two-hour conversation with a firm’s leadership team.
- AI use policy checklist: a checklist for a policy that survives contact with real work.
- Local AI bill of materials: what to buy, what to avoid, and what to measure.
- Benchmark method: how results are recorded and reported, so the numbers can be checked.
- 0001 Publish the method: why this repository is public.
What is not here
Section titled “What is not here”Client names, client data, pricing, proposals, pipeline, and anything covered by an NDA.
Related work
Section titled “Related work”- borck.consulting: the practice.
- locolabo.org: the applied local-AI lab.
- trusttool.borck.education: a free exercise on where to trust AI output.
- books.borck.education: the books, free to download.
Licence
Section titled “Licence”MIT. Use it, change it, teach with it. Attribution is welcome but not required.
Last reviewed 2026-10-31.