Notes for teams letting agents act., page 4

A SOC 2 report tells you a vendor's controls over your data were designed and, in Type 2, operated. It does not tell you what their agent will do on your work. How to read one.

A map of what sits between a language model and the system it changes: model, orchestration, tools, control, record, and the system of record itself.

Two people to override a block. Where the rule comes from, what it should cover for an agent, and why Meta's 'Agents Rule of Two' is a different rule with a confusing name.

An approval policy decides which agent actions go ahead, which wait for a person, and which never happen. Most teams have a paragraph. You need a table.

An agent control plane checks every action an AI agent wants to take against your rules before it runs, holds the hard ones for a named person, and records the result.

A permission manifest is the signed list of what an AI agent may touch, spend and decide alone. Anything not on the list is refused. What one contains, and why a prompt is not one.

Three numbers. What it costs to build, what each run costs, and what it can commit in your name. Most quotes only give you the first.

Building one agent is quick. Running it for years is the bill nobody quotes.

It works when one person you name reviews only what the system is unsure of. You set where “unsure” starts.

RPA follows steps. Agents make judgements. Your back office probably needs both.

Twelve questions that separate a system you can run from a demo you can't control. Ask us first.

An MIT NANDA report found most generative AI pilots showed no measurable return. The model was rarely the problem. The work around it was.