Personal AI memory is not company memory
Every big AI tool now remembers you. None of them remembers your company. What the difference is, and what we changed in our own setup today.
Insights
Notes from running our own companies.
What we learn putting sales follow-up, back office systems and AI to work, first in our own businesses and then in our clients'.
- 5 min read
- 3 min read
Our scheduled agents started running hours early
After a restart, some of our scheduled AI agents began firing hours ahead of time. What we paused, what we check now, and what is still open.
- 4 min read
Five places a rule dies inside an AI agent's run
Writing rules for an AI agent is the easy part. These are the five points where they stop being followed, and the check that catches each one.
- 3 min read
What a vending machine test says about agents in your back office
In Vending-Bench, AI agents running a simple business for a long stretch drifted into failure. The cause matters for anyone handing agents recurring work.
- 4 min read
CRM agents scored near zero on knowing what is confidential
A Salesforce benchmark found AI agents lose ground over multi-step CRM work and barely recognise confidential data. What a sales team should do about it.
- 4 min read
Your files hold what you wrote, not what you decided
An AI connected to your documents will find the plan you abandoned and recommend it again. Decisions and failures need a home of their own.
- 4 min read
We misread a client brief twice. Write the exceptions down
Good defaults cause expensive mistakes when nobody writes down where they stop applying. What our own misread brief taught us about rules with exceptions.
- 3 min read
The Replit database deletion, read as a memory failure
An AI agent deleted a production database during a declared code freeze. What a written company memory would have caught, and what it would not.
- 4 min read
Why your AI forgets the rules by the fortieth message
Rules typed at the top of a long AI chat fade as the conversation grows. Why it happens, and where the rules should live instead.
- 3 min read
Readers mark down anything that sounds machine-written
Customers trust copy less once they suspect AI wrote it, even when a person did. Generic output now has a sales cost. What helps, and what does not.
- 3 min read
Every enquiry got an answer. None of them got recorded.
Our AI assistant replied to every new enquiry on our messaging line and saved none of them where we could see. What went wrong, and what to test on yours.
- 3 min read
Our approval step had one flaw: you could not read the draft
We built a human approval step for AI-drafted client work. Earlier this month we found the approver could not see the whole draft. How gates go blind.