Automation and Operations
Which processes to automate first: a simple scoring method
Score each process on hours, clarity, data readiness and risk, check payback, then automate the top one or two. Worked example and pass/fail test.
Employee Lifecycle
AI Performance Review Assistant
Structured, evidence-based review drafts written from a manager's rough notes, ready for the manager to finish and sign.
Review season is when good managers stare at a blank form. This assistant takes their rough notes, goals and ratings and turns them into a structured draft in your company's format, tied to your evaluation criteria. Managers still choose the rating and every final word. They just start from a clear, fair first draft instead of nothing, and reviews across teams finally read consistently.
- Structured review drafts from a manager's rough notes and ratings
- Every draft follows your template and evaluation criteria
- Managers keep the rating and the final wording
$3,400
20–30 working days
Optional maintenance $249/mo
What it is
AI Performance Review Assistant turns a manager's rough notes, goals and own ratings into an evidence summary and a structured review draft. It makes no rating and no decision. The manager sets the rating, finishes the wording and records it in your HR system. Drafts follow your company's format across teams and review cycles.
Why businesses buy it
Performance reviews shape pay, promotion and whether people stay, yet many are rushed, vague or written in wildly different styles. Employees notice when one manager writes a page and another writes a line. This assistant gives every manager a solid, evidence-based starting point in your format, so reviews take less time and feel fairer across the company. A named approver checks drafts stay grounded in evidence. The Scope Card lists the template, criteria and output formats before you buy, so you know exactly what managers will receive.
Problems this solves
Managers put off reviews because writing them takes too long
Managers paste in rough notes, goals and competency ratings and get a structured review draft. The blank page is gone, and the manager still decides the final wording.
Reviews from different teams follow different formats
Every draft uses your company's format and evaluation criteria, guided by reference examples you supply. HR receives reviews that read consistently across teams and review cycles.
Written feedback is vague and hard for employees to act on
Drafts tie the manager's notes to your competencies and goals. The result is clearer feedback, while the rating and final say stay with the manager.
A good fit if
Your managers write reviews in wildly different lengths and styles, and reviews for similar roles end up hard to compare.
You already have a review template and evaluation criteria you want every draft to follow.
Managers tell you the blank page is the hardest part, more than deciding what they think of the work.
You have a named approver who checks that drafts stay evidence-based before they go back to managers.
Not the right fit if
You want a system that decides ratings or scores employees on its own. This assistant only drafts write-ups; ratings and final wording always stay with the manager.
You need job postings drafted rather than review write-ups. See AI Job Description Generator, which is built for that instead.
You want ongoing analysis of themes across many employees' feedback rather than one-at-a-time review drafts. See AI Employee Feedback.
You need more than 3 output formats or a second review cycle running at the same time. Add a note at checkout and we quote the extra scope, or start from our Custom AI request form.
How it works
- A manager enters rough notes on an employee's work over the period, plus the goals and competency ratings your process already uses.
- The assistant pulls from your review template and up to 20 reference examples of reviews you consider well written, so tone and structure stay consistent.
- A structured draft comes back: a summary, evidence tied to specific goals or competencies, and areas for development, laid out in your chosen format.
- The manager reviews and edits. Nothing is finalized without a person reading it; the assistant proposes language, it does not set the rating.
- The approved draft is delivered to your destination, such as your HRIS, performance platform or a shared drive, ready for the next step in your cycle.
- Patterns from approved drafts feed back into future cycles, so managers writing about similar roles or goals get a closer starting point each time.
What you provide within 72 hours
Your portal lists these the moment you order. Delivery starts the working day after everything is in and your access is checked.
- Your review template, rating scale and tone-of-voice guidance for write-ups.
- Ten to twenty existing reviews you consider well written, to use as reference examples.
- Access to the destination where finished drafts should land, such as your HRIS or a shared drive.
- The list of roles or the review cycle you want to start with.
- The name of the person who checks drafts stay evidence-based before they reach managers.
- Your business information (company basics, hours and brand voice), shared during intake.
- Admin access and API keys for every connected system, shared by adding our team as a user or creating a key you control, never by email.
- One decision-maker who can approve the template and keep the build moving.
- An API key for your own AI model account (for example OpenAI or Google Gemini), created for this build with a spending limit.
The 72-hour rule. Send everything on your checklist, including access, within 72 hours of ordering so your build starts on time.
Timeline
Start
Intake verified and your review template, criteria and reference examples confirmed.
Weeks 1–2
The drafting pipeline and evaluation-criteria mapping are built from your material.
Week 3
The destination system is connected so approved drafts land where managers work.
Weeks 3–4
25 sample review drafts are generated and reviewed against your criteria.
Week 4
Your review and two rounds of changes to structure, tone or wording.
Go-live
Live, handover report sent, support starts.
| Option | Delivery time | Price |
|---|---|---|
| Standard | 20–30 working days | $3,400 |
| Express | 15–25 working days | $4,100 |
| Rush | 10–20 working days | $4,750 |
What you receive at handover
- AI Performance Review Assistant delivering drafts to one destination (CMS, drive, store or inbox)
- One generation pipeline with up to three output formats in your review format
- Your brand profile and up to 20 reference examples set up to guide each draft
- Test report from 25 sample drafts reviewed and approved by you
- Admin guide and a written handover report
- 30 days of post-launch support
Expected benefits and how to measure them
- Can help managers finish reviews on time
- Helps reviews follow one format across teams
- Makes feedback more specific without taking the decision from managers
- Reviews submitted by the deadline
- Compare on-time completion in your review system for the cycle before launch with the cycle after.
- Manager time per review
- Ask a few managers to estimate time per review in each cycle.
- Reviews returned by HR
- Count reviews HR sends back for format or missing sections, before and after launch.
Before you start: Record on-time completion from your last review cycle. Ask managers how long each review took.
These are things to track, not promised results. Outcomes depend on your data, your team and your market.
Running costs
The build price is one-time. Running it afterwards is billed in your own accounts and typically runs about $20 to $150 a month in AI model or media-generation usage, depending on volume, mostly driven by how many reviews are drafted each cycle.
AI usage during the build and testing is also billed to your accounts, so you keep ownership of every provider relationship. Nothing runs through us.
Ask about running costsCustom requirements
If you need something outside the Scope Card, add a note at checkout. You get a written Change Quote, and that work only starts once you accept it. Your original order runs on its normal timeline either way. Larger or bespoke projects go through the Custom AI request form instead.
Request a custom versionKeep it running
Optional monthly maintenance covers monitoring, fixes, model and platform updates, a monthly health report and monthly change hours, enough for a template tweak between cycles. It starts after the 30 days of included support, bills monthly, and can be cancelled any month.
$249/mo (the minimum fee, includes 2 change hours)
Optional. Starts after the post-launch support period. Cancel any month.
How maintenance worksHow this compares
| Do it yourself with a SaaS tool | Hire a custom agency | Upmark Labs | |
|---|---|---|---|
| Price | A performance-management platform charges a per-employee monthly fee | An HR consultancy bills hourly to redesign your review process | Ours is one fixed price, shown before you buy |
| Scope | A performance platform bundles review writing into a much larger suite you may not need | A consultancy's scope depends on the engagement agreed | Ours is written on the Scope Card before purchase |
| Delivery | A performance platform requires a rollout project across your whole HR stack | A consultancy usually takes several weeks to redesign a process | Ours has a fixed delivery time shown before you buy, with faster options |
| Ratings and decisions | Some performance platforms nudge toward automated scoring | A consultancy build depends on what was agreed | Ours only drafts write-ups; the manager sets the rating |
Questions about this product
What does AI Performance Review Assistant cost to launch and run?
The build is one fixed price, shown on this page before you buy, with express and rush delivery available at a set extra cost. Running the assistant afterwards costs roughly $20 to $150 a month in AI usage, depending on how many reviews are drafted each cycle, paid directly in your own accounts. Nothing is charged per manager.
Does the assistant decide the employee's rating?
No. It drafts the written summary and ties evidence to the goals and competencies a manager already scored; the manager sets the rating and can edit any part of the draft before it goes anywhere. Final performance decisions always stay with the manager and your HR process, never with the assistant.
Will review drafts sound the same for every employee?
The structure and tone stay consistent with your template and reference examples, but the content is unique to each employee: it reflects the specific notes, goals and competency scores a manager enters. You can also approve up to three output formats, such as a short summary and a full detailed write-up.
How much of our team's time does setting up the Performance Review Assistant take?
Most HR teams spend one to two hours gathering the review template, criteria and reference examples inside the 72-hour intake window. After that, expect roughly an hour reviewing the 25 sample outputs during testing, plus the written handover report at launch.
Can we use this for multiple review cycles or role types?
Yes, within the same template and destination system. Small adjustments between cycles, such as an updated competency list, fit inside the monthly change hours included with maintenance. A second template or a parallel review cycle running at the same time is quoted as a Change Quote before work starts.
Is employee performance data kept confidential?
Yes. The assistant only processes the notes and criteria a manager enters for that draft, and admin access to connected systems is shared by adding our team as a user or creating a key you control, never by email. It does not build a separate database of employee performance history outside your existing systems.
Can managers still write reviews entirely by hand if they prefer?
Yes. The assistant produces a starting draft, but nothing stops a manager from rewriting it fully or skipping it for a particular review. It is designed to remove the blank page, not to replace a manager's judgment about their own team's work.
Is employee evaluation a high-risk AI use under the EU AI Act?
It can be, so this build keeps a person in charge. The assistant drafts an evidence summary for the manager and never sets a rating or makes a decision. The manager writes the rating and final wording in your own HR system, where it is recorded. Your HR team stays responsible for any EU AI Act duties. This page is not legal advice.
Guides on this topic
All guides Give managers a real draft instead of a blank review form.
Order AI Performance Review Assistant and your managers get structured, criteria-based drafts on a fixed delivery date, in time to plan your next cycle. The scope above is what you get, in writing, before you pay.
Works well with
Employee Lifecycle
AI HR Assistant
Answers employee HR questions from your policies and routes requests.