Clear field notes for better AI decisions.

Short, practical reads on buying, building, and running AI systems.

latest dispatchlog 0076 min read

We evaluate every vendor in 4 prompts

Compare how three models handle the four prompts we use before evaluating a vendor.

Live prompt compareBuyingRead the dispatch
log 0104 min read

Pre-launch security for solo app devs

A focused pre-launch security pass using open-source scanners with an AI review layer.

Security checklistEngineering
log 0065 min read

Should you automate this? Be honest.

Five questions to decide whether a workflow needs AI, rules, a simple script, or no automation at all.

Decision treeBuying
log 0086 min read

What a 6-month AI roadmap looks like

Six stages, what gets built when, and where costs and savings can compound.

Architecture diagramStrategy
log 0055 min read

The handoff at 60 reps, automated

When a spreadsheet handoff stops scaling, see what changed, what stayed human, and how to value the time saved.

Workflow diagramOperations
log 0093 min read

Pulse: where are you on AI adoption?

Four practical stages, the trap inside each one, and a quick browser-only poll.

Inline pollOperations
log 0035 min read

Caching is where the money is

Use the calculator to see when prompt caching reduces repeat-input cost, and when it does not.

Token cost miniEngineering
log 0025 min read

The five terms every buyer mishears about LLMs

Five LLM terms buyers hear constantly, explained in plain English and grounded in buying decisions.

Glossary widgetBuying
log 0014 min read

Cold email that reads like a person, not a template

See a real before-and-after rewrite, line by line, with the campaign context and limits included.

Before/after widgetOutreach