Fine‑tune existing Pricefx Agents after they have been created and deployed (for example by Copilot or the Agent Generator). Once an Agent is in place, you can adjust its data scope, alert conditions, action templates, and impact calculations so that it matches your specific business objectives, data structures, and governance rules without rebuilding it from scratch.
The configuration workflow is organized into four main steps:
-
Definition Step – Configure the input dataset for the Agent by choosing data sources, grouping dimensions, measures, and filters that define which records are analyzed.
-
Detection Rules Step – Define the business rules and scheduling logic that determine which records become detected items and when the Agent runs to evaluate them.
-
Action Definition Step – Specify how detected items are converted into Actions, including summaries, assignees, dashboards, grouping, and impact calculations.
-
Summary Step – Review a consolidated view of the configuration and results, including definitions, applied rules, created and not‑created actions, and expected impact metrics.
For advanced use cases, you can also draw on the shared analytical tooling used across Pricefx:
-
Expression Editor – Build and maintain custom measure formulas used in Datamart‑based Agent measures.
-
Query Expressions – Reference the full catalog of functions and operators supported in Datamart and Data Source expressions when defining measures and filters for Agents.