CEDX Timesheets · Projects & Delivery
Timesheets turns logged hours into the three numbers a services business runs on: billable share — 66% on 808h of 1,213h — leakage risk, and utilisation. Then it names the lines that leak.
The leakage board does not say “revenue at risk.” It says: 73 entries coded non-billable on client work, $58K, here are the people and projects. Specific is fixable.
Runs on demo data. Dollar figures are rate × hours on the seeded rate card, not billing data.
What it is
The board scores 184 lines and prices the expectation: value × risk% sums to $54K of expected leakage. The signals are named — “Internal time on client project”, “Missing client code”, “Rate below role / project” — with person, project, hours and value on every row.
Booked-util % against billable-util % for W31–W38 on a do-nothing baseline: average 79% across 12 people, two under 70% (bench risk), two over 85% (burnout risk), and W37 drifting to 74% before anyone calls it a problem.
The Suggested panel drafts entries from app and window titles — a document maps to the rollout project, a ticket queue maps to the support retainer — each with hours and a confidence, and each waiting on Accept, Edit or Reject. The screen says it plainly: “nothing tracked live.”
Product tour
Not a mockup and not a concept deck. This is what opens at /app/timesheets.
01 — Overview
1,213h logged, $176K billable at a 66% share, 46 sheets pending approval, time health 68. The alerts are priced: $58K of internal hours on client projects, $37K aging past the five-day approval SLA, $30K billed under the role rate card.
02 — Billable leakage
Every scored line shows its signal, person, project, hours, value, risk% and expected dollars. The top signal is coding internal time to a client project — 8.5h at $2.7K, 65% risk — and the Scoring model tab shows how the number is built.
03 — Util forecast
The forecast reads booked hours against capacity per week and per person. W31 sits at 83% booked util with 61% of it billable; W37 sags to 74% on the do-nothing baseline, which is exactly when a bench conversation is still cheap.
04 — Entries
The ledger: date, person, project, task, hours, billable flag, value, status — with chips for Billable, Leaky, Submitted, Draft and Approved. Above it, the suggested queue drafts tomorrow's entries from activity, with a human approving each one.
Who runs it
Roles, not references. We have no named customers yet, so nobody in these photographs is quoted, credited or claimed as one.
Watches the leakage board for their projects: internal time coded to client work and rate-card drift are margin conversations held in week 31, not at quarter end.
leak signals · 184
Reads the util forecast eight weeks out: two people under 70% and two over 85% is a staffing decision made while there is still time to make it.
avg util · 79%
Runs the approval queue: 46 sheets pending and a five-day SLA on the manager side, with the aging priced at $37K.
pending approve · 46The shape of it
Every figure below is legible in the captures above. Nothing here is a projection of your estate — it is the state of the demo data.
How it runs
Activity suggests the entry — app and window title mapped to a project, with hours and confidence. The person accepts, edits or rejects; nothing is tracked live and nothing auto-submits.
The sheet goes Draft → Submitted. The entry ledger keeps billable flag, value and status on every row.
Managers clear the queue against a five-day SLA — 46 are pending in the demo, and the 25 aging past the SLA are priced at $37K.
Leakage, utilisation and billable share close the loop: the week's 1,213h become a P&L reading, and the forecast projects it eight weeks forward.
One record
An approved hour is one record with many readers: the engagement's margin, the planner's utilisation, the pay run. Nobody re-keys it.
Engagement margin and unapproved time on the Services overview are these hours, read against SOW values.
CEDX ResourcingThe capacity grid's planned hours and the timesheet's actual hours are the same people's week, planned and lived.
CEDX PayrollThe payroll export tab is on the Timesheets nav because the approved hour is what the pay run should consume.
CEDX ProjectsPoints move the board; hours explain the points. The two read the same people's week.
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 rate card, your projects
Timesheets 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/timesheets. It runs on demo data.
The suggestion feature maps app and window titles to projects and drafts an entry for a human to accept, edit or reject — and the demo screen states “nothing tracked live” on itself. The entry is not filed until the person approves it.
Three named signals in the demo build: internal time coded on a client project, a missing client code, and work billed below the role's rate card. Each line is scored with a risk% and priced at value × risk% — $54K of expected leakage across 184 lines.
Eight weeks. The demo shows W31–W38 on a do-nothing baseline — booked util against billable util per week, per person — with W37 sagging to 74% and flagged before it becomes a bench problem.
They get priced. Sheets aging past the five-day SLA are summed and displayed — 25 sheets, $37K in the demo — so an approval bottleneck is a number, not a nagging feeling.
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 look up what last Friday would have scored on the leakage board.