Glossary

The terms we use, defined in plain language.

Accountable executive
The person who owns the outcome of a task and makes the final decision on how it's done. The analyst recommends; the accountable executive decides.
Expected loss
The cost of each way a task can fail, multiplied by how likely it is, added up across failure modes. Part of NET.
Freed capacity
The human time a human-AI decision returns, measured so it can be planned against and put to work.
Gottfredson-Cavalier Matrix
A matrix of task difficulty against the critical impact of failure, read with the Human-AI Task Scale. More
HMPA
Human-Machine Performance Analyst: the role that maps work, sets the human-AI split, prices it and recommends a decision. More
Human-AI Task Scale
Seven levels describing how a task is shared between a person and AI, from L1 (human only) to L7 (fully autonomous agent). More
NET
The value of a task's output compared with its full cost: human, AI, orchestration, supervision and expected loss. More
Orchestration cost
The cost of connecting people, AI tools and systems so a task runs end to end.
Supervision tax
The cost of overseeing, auditing and recovering from AI work. It grows as AI takes on more of a task, and it belongs in the price.
Talent Rewire
The company, and its licensed platform that runs all nine steps of the method on an organization's own workflows. More
Task
A unit of work inside a workflow, made of steps. Decisions about AI are made at the task level.
Task intelligence
Knowing which tasks to automate, which to augment and which to keep human. It connects goals, activities, tasks and skills.
Throughput realization
How much of AI's extra output an organization can actually use, given review and downstream capacity.
Workflow
A sequence of tasks that produces an outcome, such as drafting a credit memo. The first thing Talent Rewire documents.
ΔNPV
The net present value of moving a task from one Task Scale level to another, with every term a change from today. More