Does Vareons support an AI Copilot app? What does it show?
Last verified: 2026-09-19
The AI app in Vareons works from the AI usage page in the dashboard: it shows total tokens used, average tokens per request, active models, a prompt log with CSV export, usage distribution, efficiency insights, and a tool audit trail. Provider keys are managed from the keys page with save and test actions, per-store encryption, and masked previews in the interface. See the AI app documentation and AI key management.
Steps to follow usage
- Open the AI usage page and review total tokens and active models.
- Browse the prompt log to learn the most repeated uses and export the log when analysis is needed.
- Review usage distribution and efficiency insights to tune reliance on tools.
- Manage provider keys from the keys page, testing each key after saving.
Keys and audit
Keys are encrypted per store and appear masked in the interface, while the audit trail documents tool usage. See AI copilot overview, manage keys and usage, and the merchant copilot.
Controlling cost and usage
Follow total tokens and usage distribution weekly to catch any excess early, and steer the team toward the highest-value uses instead of random exploration. Review the audit trail regularly to confirm tool usage stays within permissions, and refresh keys on any provider-side change so work never halts suddenly.
- Set a weekly usage target and watch adherence to it.
- Disable unused tools from plan policy where possible.
- Export the log monthly and analyze patterns to improve efficiency.
Frequently Asked Questions
What does the AI app show?
It shows total tokens used, average tokens per request, active models, and a prompt log with CSV export, plus usage distribution, efficiency insights, and a tool audit trail.
How do I add a new provider key?
From the AI keys page enter the key, save it, then test it from the same page to confirm it works; it is stored encrypted for your store and shown masked.
Can the usage log be exported?
Yes, the prompt log supports CSV export, so it can be analyzed outside the platform or shared with your team to review usage patterns.