Performance Coach

The agent drafts review summaries based on collected data — goal completion rates, peer feedback quotes, manager notes, and project outcomes. The manager reviews the draft, edits it to add personal context or adjust tone, and approves it before the employee sees anything. The agent provides the evidence; the manager provides the judgment.

It pulls from OKR/goal tracking (native or integrated), peer feedback responses, manager 1-on-1 notes, project management tools (Jira, Asana, Linear), code review metrics for engineering teams, and any custom metrics you define. The more data sources connected, the more comprehensive the review drafts.

The agent collects feedback and data from all teams an employee contributes to. The primary manager sees a consolidated view with section breakdowns by team. Dotted-line managers can submit feedback that appears in the review draft alongside direct-line data.

Yes. You define the review template — sections, rating scales (1-5, exceeds/meets/below, or custom), competency frameworks, and required fields. The agent generates drafts that follow your format exactly. Templates can differ by department or role level.

The agent uses the full review period — typically one quarter or one year, depending on your cycle. It weights data evenly across the period to eliminate recency bias. If you switch to Zimyo mid-cycle, the agent uses whatever data is available from connected sources and notes the coverage gap.

After three ignored nudges on the same item, the agent escalates to the manager's manager or HR, depending on your escalation policy. The escalation includes what was due, when the nudges were sent, and the current status. This is configurable — you set the threshold and escalation path.