Skip to content
GLOBAL IETA
Global IETA / Product guides
Developer tools

AURA Foundry: prepare an AI workload and read its job status

Prepare workload inputs and completion criteria for AURA Foundry. Distinguish queued, running, simulated and completed results in an authorised job review.

By Global IETAPublished 8 Oct 20262 min read

A workload is not complete just because a request was accepted. AURA Foundry manages infrastructure and workload job status, including GPU and CPU execution paths. Access is protected, so preparation begins with an authorised operational workflow.

Define the expected output

Write down the task, permitted input material and the result that would count as complete. Identify which output files or records need checking and who will review them. Keep credentials and secrets out of the workload description.

For a small initial run, use a bounded input and an observable completion criterion. This makes it easier to determine whether the execution path actually produced the requested result before attempting a larger job.

Read job states as separate stages

  • Accepted or queued: the system recorded the request; output is not yet established.
  • Running: an execution path is working on the request.
  • Failed: inspect the available error and responsible operational route.
  • Completed: check the actual output against the original acceptance criteria.

A public homepage or status label is not a guarantee that a particular workload is ready to run. Available execution paths depend on the authorised environment and connected capacity.

Keep production evidence separate from simulations

A simulated result can help explore a flow, but it must not be counted as a completed production workload. Verify the native job record and output, and keep the job reference, input version and review result together.

If the output is incomplete or does not meet the brief, record that limitation rather than treating the status alone as success. Only authorised users should inspect operational details or initiate supported workloads. A small, traceable first run provides a better basis for the next infrastructure decision than an unverified capacity assumption.

Explore AURA Foundry

Visit the official AURA Foundry product page for current access and available features. Use Global IETA product guides to explore the connected ecosystem.

About this guide

AI-assisted editorial content from Global IETA, checked against our product information on 2026-10-08. Features and access can change; check the official product page for current availability.

AURASAME CONVERSATION · EVERY SITE