Reviews that draft themselves
from the work that actually happened.

The AI assembles the evidence — OKR progress, 360° themes, prior commitments — and proposes phrasing the manager can use or rewrite. Nothing leaves the review window without an explicit manager save. Every sentence the AI proposes carries a citation back to its source data, so reviewers can audit why a claim was made.

Standard OKRs — Objectives (what you want to achieve) with measurable Key Results — implemented at company, unit, team, and individual levels. The visual builder handles the cascade and parent/child linking; weighted progress rolls up automatically. We don't impose a methodology beyond OKR conventions, so you can adapt it to your house style.

Cycles can be annual, quarterly, monthly, or custom. They nest — an annual cycle can contain four quarterly ones with their own KRs and check-ins. Each cycle carries its own evaluation rules, scheduling, and rater sets. Transitions between cycles are automated, with reminders for unfinished evaluations.

Yes. Employees see their personal dashboard with goal progress, prior reviews, and development plans. Once a manager releases the evaluation, the employee can acknowledge it and add a written response that becomes part of the cycle record.

Goals carry weights expressing relative importance. Progress is a weighted average of KR completion, factoring in dependencies. The dashboards show both raw progress and risk-adjusted forecasts on the in-progress cycle — so a 50% KR with a known blocker doesn't read the same as a 50% KR on a healthy trajectory.

Yes. Build different forms for different roles or departments. Define your own competencies and rating scales — Likert, percentage, custom anchors. The AI drafting layer adapts its phrasing to the form's structure, so a 5-anchor behavioral form gets different language than a percentile-ranking comp committee form.

Goal completion rates, evaluation distribution across rating tiers, cycle-over-cycle trends, team and department breakdowns, calibration drift, KPI tracking. Exportable as CSV, PDF, or scheduled to land in inboxes. Custom reports via the analytics module for People Analytics teams.

Access enforced at retrieval — managers see their direct reports, skip-levels see two layers, calibration committees see the cohorts they're calibrating. 360° anonymous responses are stored with rater identity stripped at write time, not just hidden in the UI. Full audit log on every read and edit, exportable for SOC 2 review.