Actual month-to-date Azure bill for this app's resource group (all services), read live from Azure Cost Management. This is the true infrastructure cost — separate from the per-user AI model usage below. Azure figures settle ~1 day behind.

Requests that tripped a security-policy rule — e.g. trying to have the AI log in with credentials, hardcode a secret, or bypass a control. The AI still declined; this is for coaching/visibility.

Crawls the W-Training source-of-truth site (manager-only & system libraries excluded) into Azure AI Search. Runs in the background.

When on, each person can choose to let Wisenbaker AI remember their role and preferences across chats (off for each person until they turn it on). Admins can never see what is remembered. Turning memory off deletes everyone's memory topics; preferences are kept.

Available to every user, off by default. Each person chooses which company skills to turn on in Skills. New skills show a New indicator until that person opens Skills. Pausing a skill here makes it unavailable to everyone.

Suggested by staff

Skills people sent in for the whole company. Review opens the text for you to edit; Publish makes it a company skill (still off for each person until they turn it on).

Real questions people ask, put through the app exactly as a person would, with the right answer read straight from Echelon for the same dates. A run takes a few minutes and costs about the same as asking the questions yourself (it is billed to you). Nightly: runs as you at 3 am; if something that passed before now fails, you get a Teams message at 7 am.

👍 / 👎 users gave on answers. Use the 👎 notes to fix prompts, SOPs, or the index.

Activity by user (click to expand): budget changes, re-index, email drafts, and 👍/👎 feedback. Append-only. To keep costs low we don't log every message — only these events.