AI usage and observability

Know where the intelligence is going.

Understand which work uses your AI capacity and where it needs attention. Cedros brings usage totals, cost estimates, provider and model breakdowns, and recorded failures into one view, so you can make better decisions about how the assistant works.

Explore plans /pricing See the platform /about

Understand the workload before changing it

Understand usage over time

Review recorded AI calls, tokens, success and failure counts, and estimated cost for a selected period. Daily totals show how the workload changes, while month-to-date spend provides context for your configured spending guard. Start with the overall pattern before investigating individual requests.

Find the work behind the total

Break usage down by feature, model, provider, routine, or job. Inspect the most expensive recorded calls to see where a deeper review may help. A recurring content task and customer support may benefit from different model choices or frequencies.

Investigate failures with context

Review recorded outcomes and provider readiness when work is slow or unsuccessful. Use the nearby activity history and operational logs to follow what happened. This makes it easier to identify whether the next step is adjusting a task, checking setup, or investigating a provider issue.

Make informed capacity choices

Compare the work being done with the capacity it consumes. Change model selection, routine frequency, or spending settings in the relevant configuration area. The usage view helps you assess those decisions while keeping the figures tied to recorded activity.

Understand what the cost estimate includes

Cost estimates use recorded token usage and known model pricing. Calls without a known price are identified as unpriced rather than counted as free. Use provider bills or managed-plan records for the final amount charged, especially when reviewing a period with unpriced calls.

Keep activity and delivery health close by

The Observability workspace also includes site activity, recent operational logs, and email delivery health for authorized users. Open these views when you need to understand a change or a delivery issue, then return to the relevant feature to make adjustments.

Useful to know

Can I see which routine is using the most AI?

Yes. Usage reports include routine and job breakdowns alongside feature, provider, and model totals. The report also highlights expensive individual calls.

Are estimated costs the same as my invoice?

They are estimates based on reported usage and known pricing. Unpriced calls are flagged, and provider or managed-plan billing remains the reference for the final charge.

Can I change budgets from the usage report?

The report shows spend and relevant guard or readiness information. Budget, provider, and model changes belong to their configuration settings.

Put useful intelligence on a sustainable footing.

Review the work, understand its usage, and set boundaries that fit the business.

Talk about your website /contact