Notes from the automation frontier
Playbooks, research and honest lessons from putting AI agents into production for real teams.
Latest notes
Published every other week
Playbooks
7 min read
How to pick your first AI automation (and not regret it)
The best first project is boring, frequent and measurable. Here’s the scoring framework we use with every client.
Playbooks
6 min read
Human-in-the-loop, done right
Approval steps can make agents safe — or make them useless. Five patterns for oversight that doesn’t slow you down.
Research
9 min read
What 40 AI deployments taught us about ROI
We looked back at every agent we’ve shipped. The biggest returns didn’t come from where we expected.
Product
8 min read
Prompt contracts: instructions agents actually follow
Long, clever prompts are fragile. Short, testable contracts are not. How we write instructions for production agents.
Product
6 min read
Inside our evaluation stack
How we know an agent is ready for production — and how we know when it stops being ready.
Opinion
10 min read
From chatbot to coworker: the state of AI operations in 2026
Agents have moved from answering questions to doing work. What changed, and what still doesn’t work.