CEDX Performance · People & HR
Performance runs the whole cycle — goals, check-ins, reviews, calibration — and scores its own health: completion risk per packet, ratings inflation per calibration group, and a distribution that prints what each band is for.
The rating distribution prints its own rule under the bars: 1–2 need action plans; 4–5 feed promotion and retention lists. A scale that says what it is for is a scale people can argue about honestly.
Runs on demo data. The cycle names itself H2 2026 Merit on screen.
What it is
Three hundred twenty review packets across Not started, Self review, Peer review, Manager review, Calibration, Shared and Complete — with rating, goals progress, due date and manager per row. Forty-four carry a completion-risk score of 50 or higher.
Two hundred ten goals typed OKR, Individual, Team or Development — “Enterprise pipeline $4.2M”, “Stabilize on-call”, “Document runbooks” — each with progress and weight. Statuses include Dropped, because some goals should die in the open.
One hundred eighty weekly check-ins over three weeks: 44 submitted, 94 late or missed, average pulse 3.1 on a 1–5 scale. The late column counts days, so a missed week is a nudge, not a surprise at review time.
Product tour
Not a mockup and not a concept deck. This is what opens at /app/performance.
01 — Overview
Cycle health sits at 42 — built from 26% complete, 44 high completion risk, 5 bias flags, 53 packets in calibration and a 3.5 average rating. Alerts name the risk: a self review unstarted with the deadline inside a week; calibration inflation at mean 4.04 against an org 3.5.
02 — Reviews
Search by employee, department or manager across 320 packets. The distribution panel reads 0, 18, 70, 68, 14 across the five bands and prints what the bands mean — action plans at the bottom, promotion and retention lists at the top.
03 — Goals
Two hundred ten goals sorted by progress, from 99% down to the stalled. Type chips split OKR, Individual, Team and Development; status runs On track, Behind, At risk, Dropped and Complete.
04 — Check-ins
One hundred eighty check-ins with week, status, pulse, priority, manager and days late per row. Submitted, Late, Missed and Draft are tabs, so the 94 needing a nudge are one click from a list.
Who runs it
Roles, not references. We have no named customers yet, so nobody in these photographs is quoted, credited or claimed as one.
Writes the self review and the weekly check-in — and sees goals progress on the same row as the rating.
pulse 3.1 · goals 50%
Watches cycle health: completion risk, the 53 packets in calibration, and the bias flags before they become arguments.
health 42 · 5 bias flags
Runs the room where mean 4.04 against org 3.5 is a slide, not a whisper — inflation is a number on screen.
53 packets · mean 4.04 vs 3.5The shape of it
Every figure below is legible in the captures above. Nothing here is a projection of your cycle — it is the state of the demo data.
How it runs
Goals are typed and weighted at the start of the half — OKR, Individual, Team, Development — 210 of them in the current cycle.
Weekly pulses keep the year honest: 3.1 average across 180 check-ins, with the late and missed counted by day.
Packets move self → peer → manager, each with a completion-risk score, so a stalled review is chased before the deadline, not after.
Fifty-three packets sit in calibration with bias flags watching inflation — mean 4.04 against an org-wide 3.5 is visible to the whole room.
One record
Goals, learning paths, focus data and the roster are the same people — the cycle sits on top of them.
The 180-review H2 2026 cycle appears inside the People captures too — same roster, same record.
CEDX LearningDevelopment goals point at paths: the academy is where a growth goal becomes coursework.
CEDX Recruiting“Hire for team” is a live goal in the captures; open reqs and review packets share the record.
CEDX TimeFocus and fragment data is the context a check-in conversation sits on.
Limits
Finding this out on the third call is worse for you than reading it here, and worse for us.
Start
Open it right now
Your cycle, your bands
Performance with the rest of it
Questions
Yes. Every screenshot is a capture of the running build and you can open the same build at /app/performance. It runs on demo data, which the cycle labels on screen.
A score per review packet that flags work likely to miss the cycle — 44 packets at 50 or higher in the demo, with the stage and score on the risk board. The top alert is a self review unstarted with the deadline inside a week.
Calibration drift, shown as numbers: the demo alert is a group whose mean rating is 4.04 against an org-wide 3.5. It is an analytics flag, not a fairness certification.
Check-ins are the weekly pulse — submitted, late, missed, with a 1–5 score — and reviews are the cycle packets. In the demo, 94 late or missed check-ins are the nudge list that keeps the cycle honest.
No certification has been issued. What we can evidence about hosting, encryption, tenant isolation and production access is written up on the security page.
Live build, demo data, no card. Then check how many of your own reviews would be flagged at risk today.