I want to help local businesses use AI
The useful work is rarely a chatbot. It is finding a repeated process with enough volume, earning permission to touch it, pricing the risk honestly, and leaving behind a system the business can operate without you.
You end with: One scoped, priced and measured engagement that a real business can keep running after you leave.
Explore
Learn to reject impressive ideas whose volume and value cannot repay the work. A good audit produces more noes than proposals.
- The economics of automation — Why one job gets automated and the near-identical one next to it does not. (3 lessons, 65 min)
Build
You need one fast build, evidence for its exception rate, and a deployment option that respects the data boundary of a small business.
- Coffee break: the summariser you can prove is wrong — Twenty minutes to build it, twenty more to find out where it fails. The second half is the point. (2 lessons, 40 min)
- The eval you actually need — Stop reading leaderboards. Build the harness that measures your task, on your inputs. (4 lessons, 8 min)
- Local or nothing — Run open-weight models on hardware you own, and know what they cost you in watts and seconds. (5 lessons, 6 min)
Earn
Test whether the problem is real before proposing a solution, then write a quote whose exclusions and running costs are visible.
- Validate it in 48 hours — Find out whether anybody wants it before you spend six weekends finding out they don't. (3 lessons, 65 min)
- Price the work — Scope it, cost it, quote it — and know the number below which the job loses you money. (3 lessons, 80 min)
Launch
Do the whole engagement inside somebody else's business: permission, baseline, deployment, measurement and a clean handover.
- Automate a local business — Somebody else's business, somebody else's process, and a handover where you never get called again. (23 lessons, 1010 min)