Business workflows

Nobody buys features. You buy the ability to ship, support, and improve connected devices as a business.

Most pages are short stories: recognisable moments for people who build and run Linux IoT devices - either as an internal capability (sites, plant, field assets) or as the product they sell. Each story shows what usually goes wrong, how Dataplicity fits the workflow, and where to go for how-tos. Platform depth provides longer guides to capabilities that become more powerful when they work together.

Feature reference stays in Fleet, Operations (including customer surfaces), and Developers. Stories live here.

Stories and screenshots often name illustrative customers - Northstar, Harbor, Polar, Solstice, Amber Field, and Meridian. They are demonstration fleets that make the operating patterns concrete. The methods are real; the companies are illustrative. For a longer screenshot-led OEM tour, see SiloSentry (also fictional) and the screenshot tour.

Choose a path

Path	Who it is for	Start here
Illustrative customers	Recognise your business in the demonstration fleets	Northstar, Harbor, Polar, Solstice, Amber Field, Meridian
Internal IoT operations	Devices that support your business	Fix a remote site without a truck roll
Connected product company	The device is what you sell	Ship the first customer unit you can still support
Field advantages	Network, hardware, software-age, and on-call realities in deployed fleets	Starlink vs product, mixed-age images, phone on-call
Crafted surfaces	Dashboards, wallboards, and status done with intent	Design the wallboard for who walks past it
Platform depth	Operators evaluating how the platform handles evidence, cohorts, control, and poor connectivity	Logging, browser workspace, fleet monitors, and file requests

Platform thesis

New to the post-shipment framing, or coming from remote shell only? Read After devices ship.

Platform fit

Already have a runtime (Greengrass, balena, systemd, containers, your own supervisor)? Start with Keep your runtime; add the operations layer.

Start from a moment

Business moment	Story
First site device into a known state	Get the first site device into a known state
Dark assets before the shift	Know which assets are dark before the shift starts
Remote break/fix without dispatch	Fix a remote site without a truck roll
Contractors without master keys	Let contractors help without handing them the keys
Prove maintenance actually ran	Prove what changed after a maintenance window
First commercial ship	Ship the first customer unit you can still support
Factory identity to field ticket	Carry one device record from factory to field ticket
Engineering still is tier one	Hand support from engineering to a real queue
Field support escalation	Decide dispatch vs remote fix on a field support call
Customer wants truth without the shell	Tell the customer what is true without the shell
One unit vs the network died	Separate one bad unit from the network died
Evidence for dispatch and access	Keep evidence for field decisions and who did what
Fleet looks half offline	Blame Starlink (or not) before you blame your product
On-call away from a desk	Take the incident on the phone without losing the plot
Old and new Linux images together	Keep one ops layer across ancient and brand-new images
Wallboard / TV that nobody trusts	Design the wallboard for who walks past it
“Does this replace our runtime?”	Keep your runtime; add the operations layer
Device output is scattered across tools	Logging for a fleet, not a terminal
A flat offline count hides the affected population	Monitor the cohort, not just the device
A field device will not stay online for support	Request the evidence now; let connectivity catch up
“Is the web app more than a remote terminal?”	Run the fleet from one browser workspace
See all four capabilities in one response	Diagnose the Northstar cold-chain incident