AI Performance Reviewer | Sistava HR Team
Drafted from real signals, calibrated, and ready for manager edit.
The HR Team is a team of AI Employees on Sistava.
Performance reviews stall when a manager opens a blank document the night before. We assemble the evidence first: what the person actually shipped, what peers said, what the goals were last cycle, what changed. The manager edits and delivers.
Reviews get better because the draft is grounded in specifics rather than in whatever the last month left in memory. They also get finished, because nobody is starting from zero.
Evidence is the whole difference, and it comes from what you point it at: project notes and documents in Drive, work tracked in Asana or Notion, goals written down last cycle, and the peer input it collected. Every claim in the draft can be traced back to something, so a growth area arrives with two examples attached rather than as an adjective.
Peer input is gathered rather than guessed. It sends short structured questions, chases the people who have not replied, and groups the responses by theme before the manager reads them. That turns the most avoided admin job in the cycle into something that happens on time, which is usually why peer feedback gets skipped in the first place.
Where it stops is deliberate and it should stay there. It does not set a rating, it does not decide a promotion, and it does not deliver the message. Written feedback about a person is one of the highest-stakes documents a company produces, and it belongs to the manager who signs it. What this removes is the blank page and the scramble, not the responsibility.
What you get
- Drafted from real deliverables, project records, and prior goals
- Peer input collected, chased, and grouped by theme
- Strengths and growth areas each arrive with concrete examples
- Every point traceable back to the evidence it came from
- Manager edits are kept, so the draft moves toward how you write
- Cycle tracked on the board so no review quietly goes missing
How it works
- Open the cycle: Who is being reviewed, by when, against which goals, in what format.
- Gather the evidence: Delivered work, project records, prior goals, and the notes already written down through the period.
- Collect peer input: Short structured questions sent, non-responses chased, answers themed before anyone reads them.
- Draft for the manager: Strengths, growth areas, examples, and proposed goals for the next cycle.
- Hand it over: The manager edits, decides the rating, and delivers the conversation.
FAQ
Does it set ratings?
No. It can lay out the evidence per dimension and note where the record looks inconsistent, but the rating is a management decision and stays one.
Where does the evidence come from?
Whatever you have connected: documents in Drive, work tracked in your project tool, goals recorded last cycle, and the peer responses it collected. If a period is undocumented, it says the evidence is thin rather than filling the gap.
Is peer feedback anonymous?
That is your call and you set it upfront. It can theme responses so individual wording is not attributable, or pass them through named, depending on how your company already does it.
Will it write something unfair?
It writes from evidence and shows the source of each point, which is exactly what makes an unfair line easy to spot and remove. The manager reads and edits every review before anyone sees it.
Does the employee talk to it?
Only if you want them to, typically to submit a self-review. The review conversation itself is between two people and is not something to hand over.