Notes on Agents That Actually Do Things
For people deciding whether an agent is the right fix for a task, and what it would take to build one properly. Specific and a little wry about AI hype, not another "10 ways AI will change your business" list.
Nothing Published Here Yet
This is a new site and the blog is genuinely empty right now. We'd rather say that plainly than fill the page with placeholder headlines that link nowhere, or posts written to hold a keyword instead of to say something useful.
When we do publish, it will cover the same ground we work through on discovery calls — how to tell whether a task actually needs an agent, what makes a build simple versus complex, and where the honest limits are. Until then, the pages linked below already cover most of it in more depth than a blog post would.
Subjects We Keep Coming Back To
Not a general AI news feed — the specific questions that decide whether an agent build goes well.
Agent vs. Chatbot vs. Automation
Where the lines actually sit, and why the label on a tool matters less than what it's allowed to do.
Scoping & Escalation Design
What a good brief includes, and how to decide where an agent should stop and hand off to a person.
Integrations in Practice
What actually makes a system easy or hard to connect an agent to, beyond "does it have an API".
Cost & Timelines, Honestly
What actually drives an agent's price and build time up or down, without a padded number attached.
Skip the Reading, Bring the Task
A short discovery call is faster than working through a reading list. Tell us what the task looks like today.