Some questions end at a figure. Many carry on: why it moved, how it compares with last year, where it lands next week, and who should read it on Monday morning.
Four causes together account for 95% of the fall, and supplier confirmations moving later is the largest.
Four steps; each one rests on the rows a reader can open.
Clear to build is the share of a build plan that has every component on hand. When it falls, the useful question is why, and a useful answer names causes a person can act on this week.
Juno ranks candidate causes by how much of the change each one accounts for and attaches the rows behind each candidate. A reader agrees or disagrees with the evidence on the same screen as the claim.
Each candidate carries a confidence. Above the floor your team sets, the candidate appears with its rows. Below the floor, Juno says the evidence is thin and leaves the suggestion open, which is the honest answer more often than people expect.
Forecasts follow the same habit. A short forecast arrives as a range with its basis named, and Juno compares it with what happened, so the accuracy record sits beside the forecasts themselves.
Six steps; the arithmetic is simple and every candidate keeps its rows.
The method is plain arithmetic over your own rows. Juno holds each input of the measure at its earlier value in turn and works the measure out again. The gap that opens is the effect of that input alone.
That keeps the suggestion checkable. A planner who disagrees opens the 18 components whose confirmed dates moved and reads them, one line at a time.
Below the confidence floor, Juno states that the evidence is thin and names what would settle it: more weeks of history, a source carrying inspection results, or a confirmed mapping for one column.
A reader can add a cause of their own with a sentence. The addition joins the record and teaches the next ranking.
The measure definition travels with the answer, so two people reading the same cause read the same arithmetic.
History gives the shape; the range gives the week ahead, and both are checked.
| Measure | The range forecast a week ahead | What happened | Inside the range |
|---|---|---|---|
| Clear to build, Penang, week 39 | 76% to 79% | 78.4% | yes |
| On-time in full, week 39 | 93.2% to 94.6% | 94.1% | yes |
| Days of supply, 7741-X at 1210 | 19 to 23 days | 24 days | one day above |
| Late lines, Delta Fasteners | 7 to 12 lines | 9 lines | yes |
A comparison reaches as far back as the registered sources hold. Where a source keeps a short history, Quanta lands a governed copy in your lake and the comparison reads it there, under the same access rules.
A short forecast covers a few weeks, over the same rows the answers use: open orders, confirmed dates, the published plan and the recent record of each supplier.
Ranges hold more information than a single figure. A range of 76% to 79% tells a director that the week is steady; a range of 64% to 88% tells them to look at the inputs first.
Every range is compared with the actual when the week closes. The table above is that record, and a reader who wants to trust next week’s range reads last week’s.
One saved line of questions reaches five readers, each with their own rows.
| What moved | This week | Against last week | What Juno suggests reading next |
|---|---|---|---|
| On-time in full | 94.1% | +0.7 pts | Northbay Logistics holds 36 of the 84 late lines |
| Clear to build, Penang | 78.4% | +1.1 pts | inside the range forecast last Monday |
| Days of supply, controllers | 24 days | +3 days | one day above the range forecast last Monday |
| Open lines past due | 212 | 31 fewer | Delta Fasteners carries 36 of them |
| Excess above six months | $6.40M | $0.21M more | a transfer to Dresden would move $349,300 |
A digest is a walkthrough on a schedule. The questions stay fixed, the figures move, and Juno adds the comparison and the line worth reading next. J. Ridley receives the company view at 07:00 on Monday.
Each reader receives their own. A plant manager reads their plant, a buyer reads their categories, a finance lead reads value by business unit, because every digest runs with the access of the person it is addressed to.
Alerts work the same way between digests. A rule stated in plain language watches one measure, and the alert reaches the person who owns it with the rows attached.
A digest is the point where ChatLima meets the working week: the reader opens the item that moved, asks a follow-up in the same thread, and keeps going.
Showcase estate.
Four weeks after the first questions, the Monday digest reached J. Ridley at 07:00. Clear to build at Penang stood at 78.4%, inside the 76% to 79% range Juno had forecast the week before, and on-time in full stood at 94.1%.
Two lines asked to be opened. Days of supply for the controllers came in one day above its range, and excess above six months had risen by $0.21M. J. Ridley opened the second line, read the three items Dresden had requisitions for, and sent the transfer question to L. Tan before the morning meeting.
The ChatLima screen, Teams and Slack, SmartApps, the API and the reporting connector, with one set of access rules and one record across all of them.
As far as the registered sources hold. Where a source keeps a short history, Quanta lands a governed copy in your lake and the comparison reads it there under the same access rules.
It is a short look ahead over the same registered rows, useful between planning runs. The published plan stays the plan, and every forecast names the inputs it read.
Juno says so and leaves the suggestion open, naming what would settle it: more weeks of history, a source carrying inspection results, or a confirmed mapping for one column.
Yes. Each range is compared with what happened when the period closes, and the record per measure is open to read. That record also tunes the next range.
Anyone with a Juno login. Each digest runs with the reader’s own access, so a plant manager reads their plant and a director reads the company.
Each one builds on the one before it.
Reads your rows in place.
Connect and move ↗One record per real thing.
Master and govern ↗Apps for every decision.
Operate and decide ↗Asks and acts, at your level.
Ask and act ↗Bring two systems and one decision your planners make every week. In the first week we register the data, confirm the mapping with your team, answer your questions in ChatLima and run the first agent at the level you choose.
