Live data
Screens that update as work happens: the shop floor count, the delivery board, the counter. For the handful of numbers where a nightly refresh is too late.
When minutes matter
The shop floor
Units built this shift against target, on a screen the crew can see, updated as each unit scans through.
Deliveries and dispatch
Which trucks are out, what is on them and what came back, as the drivers scan, not when the paperwork lands.
Counter sales and stock
Today's sales and what is left on the shelf, as the register rings, so the reorder goes out before the shelf is empty.
What it takes
A source that reports events as they happen: a scanner, a point-of-sale system, a machine, or a system with an API that answers in seconds. The feed carries each event into a small pipeline, and the screen reads from that.
Spreadsheets and nightly exports cannot be live, and QuickBooks updates on its own schedule. Those numbers stay on the dashboard, refreshed hourly or nightly, next to the live ones.
How it runs
Each event lands in a queue, a small program checks it and writes it to the database, and the screen updates within seconds. Alerts go out when a number crosses a line you set.
It runs on managed infrastructure billed at cost, inside a data team plan that watches it.
What it costs
From $12,000, quoted fixed after a paid scoping step that confirms the source can feed events. The streaming infrastructure is billed at cost, with no markup.
Most businesses need live data for two or three numbers, not the whole dashboard. Scoping finds which ones. Every price is on one page.
Before we start
What we ask
- The event: a scan, a sale, a machine cycle, a delivery
- Where it originates: a device, a point-of-sale system, a system with an API
- What should happen when it arrives: a screen update, an alert, both
- How fast is fast enough: seconds, or minutes
- How many events an hour, on the busiest day
- Who watches the screen, and where it hangs
What we need from you
- A source that reports events as they happen, or an API that answers in seconds
- Someone who can confirm the event volume
- A decision on what the alert threshold is
- A data team plan to watch the pipeline after launch
What will not work
- A source that only exports nightly
- A spreadsheet as a live source
- QuickBooks as a real-time feed (it updates on its own schedule)
- Live numbers nobody acts on