AI performance review support that runs all year, not once a year
Feedback and goals are collected as work happens, so the review is written from facts and not from memory.
Reviews are written from memory, once a year.
Managers try to remember twelve months of work in one evening, recent events weigh more than anything else, and the development plan is filed and forgotten until next year.
- HR leads who run review cycles and want them to mean something
- Managers with large teams and little time to write reviews
- Companies using OKRs or quarterly goals that slip mid-quarter
The agents
Runs the review cycle in Lattice or Google Forms, collects peer and self feedback, and follows up on development plans each month.
Never talks to anyone outside the company. Reminds managers and peers of their deadlines in Slack and makes sure every review has the input it needs.
Drafts each review from the goals, feedback and notes collected, using specific examples, for the manager to edit.
How it works
- Goals and OKRs are set in Lattice or Notion at the start of the period.
- Each month, Bia asks for short check-ins and peer feedback in Slack or Google Forms.
- At review time, Mat collects what is missing and reminds those who are late.
- Dai drafts the review with examples, and the manager edits and approves it before anyone sees it.
- Bia turns the agreed development plan into monthly check-ins until the next cycle.
Example of what you get
Review draft for a manager
- Goals and OKR progress with dates and evidence
- Peer and self feedback, grouped by theme
- Strengths and areas to grow, each with a concrete example
- Proposed development plan with three actions
What changes
- Reviews are based on a year of facts, not the last month
- Managers spend their time on the conversation, not on writing from scratch
- Development plans are followed up after the review
Where a person is still needed. The team does not rate people, decide raises or promotions, or send a review without the manager's approval. Those decisions belong to managers and HR.
Connects to
Performance review team
Is an AI performance review generator fair?
It drafts from the evidence collected during the year: goals, feedback and notes. The manager reads, edits and owns the final text. Writing from facts helps reduce recency bias, but judgment stays human.
Can AI employee feedback be anonymous?
Yes, if you set it that way. Peer feedback can be shown grouped and without names. Each company decides its own rules.
How does an AI OKR assistant help?
It checks progress on goals each month, asks owners for updates in Slack and points out goals at risk early. AI for performance reviews works better when goals are tracked all year.
Is review data private, and how long does setup take?
Reviews stay in Lattice, Notion or your forms, visible only to the people you allow, and every action is logged. Connecting the tools and your review template takes a working session, and opening the team is free.
How does AI help with performance reviews during the year?
Bia asks for short monthly check-ins and peer feedback in Slack or Google Forms, so evidence builds up as work happens. At review time, Dai drafts from that record instead of from memory, and the manager edits and approves.
Which tools does the review team connect to?
Lattice or Google Forms for the review cycle and feedback, Lattice or Notion for goals and OKRs, and Slack for reminders and check-ins. Reviews stay in those tools.
Can managers change the tone and content of the draft?
Yes, the draft is only a starting point. Managers edit every word, remove what they disagree with and add their own view. Nobody sees the review until the manager approves it.
What happens when there is not enough feedback for someone?
Mat reminds the peers and the manager who are late, and the draft says clearly where evidence is missing. Dai does not fill gaps with generic phrases.
Can we use our own review template and competencies?
Yes. Share your template, rating questions and competencies, and the drafts follow that structure. The team does not rate people; ratings stay with the manager.
Is an AI review assistant the same as an AI that evaluates employees?
No. It collects feedback, organizes evidence and drafts text for the manager. It does not score people or decide raises and promotions, which stay with managers and HR.