Runtime in Creatio AI Studio
When AI Agents serve real users, you need to release changes to them safely, see what they are doing, keep their consumption within predictable limits, and account for every action after the fact. The Runtime section of Creatio AI Studio groups the sections that cover this operational side. They control where agents run, monitor how they execute, meter what they consume, and record what happens on the platform.
Environments and deployments
Agents do not serve end users straight from the designer. As a creator, you first publish and deploy an agent version, then promote it forward as it proves itself. By default, every workspace has three environments: Sandbox, Stage, and Production. Sandbox is the entry environment, the only one that accepts a direct deploy. Stage and Production receive a version by promotion, one step at a time. The Environments app lists these environments, plus any custom ones you add, for your workspace. The Deployments app manages the rollouts: publish and deploy a version to Sandbox, promote it to Stage, then to Production once it proves ready, and roll it back if something goes wrong. This lets you test agent changes safely, without touching the version your end users rely on.
Observability
When an agent fails or behaves unexpectedly, this is where you find out why instead of guessing. Each request an agent processes is recorded as a run. The Observability app shows runs across all agents in one place: which are running now, which are completed, and which need attention. Open a run to inspect its execution step by step: the input it received, the output it produced, the tool and LLM calls it made, and the policy events that fired along the way.
Usage and budgets
The Usage app meters what your agents consume. Creatio meters consumption in AI actions or in tokens, depending on your workspace configuration. The dashboard shows consumption over time and breaks it down by agent and by model, so you can tell steady growth apart from a sudden anomaly and identify the agent behind it. It also tracks utilization against each budget you configure and warns you once usage reaches the warning threshold, before agents are stopped rather than after.
A budget caps how much agents can consume within a time window before Creatio warns you or blocks further requests. You can set a budget for the whole workspace, and an agent budget on top of it for an individual agent, so a single high-volume agent can run against its own limit without loosening the rules for the rest of the workspace.
Learn more: Set a usage budget or spending limit.
Audit log
The Audit Log app records platform activity and changes: who performed an action, when, through which channel, and whether it succeeded. It complements Observability. Runs tell you what an agent did, the audit log tells you what people and services did to the platform. Use it to support compliance reviews and to reconstruct the sequence of changes behind an incident.
