Techlify
A control room for 437 alarm sites in one browser tab. You can arm, diagnose and schedule across every site at once, or drop into a single building, without ever changing tool.
At a glance
Section 1 of 4
The idea
An alarm panel is a solved problem for one building: a keypad by the door, a code, four buttons and a beep.
None of that survives 437 sites across Berlin, Amsterdam, Madrid, Planegg and Hopsten. The person responsible sits at a desk in one of those cities, or none. The question stops being "is this building armed" and becomes "which of my four hundred buildings is not".
Section 2 of 4
The problem it solves
Three things pull against each other
1 site vs 400 sites
Build for the estate and one site becomes unreachable. Build for the site and the estate becomes a spreadsheet.
Two kinds of user
Someone typing a code at 6am and someone managing permissions from a laptop share only the word.
No two countries alike
A schedule that is right in Germany is wrong in Spain, and the time zone is the least of it.
One tool, two altitudes
The scope switcher sits at the top of the sidebar and everything below obeys it. Pick a single site and Dashboard becomes Site Overview: same five sections, now about one building.
The estate, made actionable
Four site states, one colour each, everywhere
Set
120 sites armed
Unset
115 sites disarmed
Partial-set
115 sites partly armed
Offline
5 sites not reporting
Every count opens the sites behind it, and in that list the states are buttons. You can arm a site straight from the list of unarmed ones.
Device states, by how urgent they are
Alarm (tamper)
A job for right now
Trouble
Low battery or no signal: next week
Omitted
Someone chose this. Show the decision
Ready
Nothing to do
Section 3 of 4
The process
Research
How multi-site estates are run, and what the panel hardware already exposes. The states the product can show were never a free choice.
The four states
User interviews
Separating the people who arm a building from a keypad from the people who run four hundred of them from a desk.
Panel vs platform
Wireframes
Every screen here comes from a folder labelled concept 2, so at least one direction came before it.
Concept 2
Testing
Tested against the questions estate managers actually arrive with.
Concept 2 chosen
Design
One status vocabulary: the same four colours mean the same thing on a pin, a row, a chip and a button.
One status system
Development
Specified across five sections, two user models and several countries, with empty, loading and long-name states drawn.
Dev-ready spec
Section 4 of 4
What I designed
Five sections carry the product: Dashboard, Logs & Reports, Devices, Users & Permissions and Automations. Keeping it to five was most of the work.
The sidebar gets out of the way
Once you know where things are, the sidebar just takes room. Collapsed to icons, it hands its width back to the data, and on the dashboard that width brings back the alarm times.
Two populations, two models
Two kinds of user, kept apart
Panel users
- Arm and disarm buildings
- Have a code, a set of sites and a time window
- Access expires on a date
- Granted in bulk by shift profile
Platform users
- Log into this web app
- Manage sites, devices and people
- Never touch a keypad
- Need permissions only
Putting both in one list behind a role dropdown was the obvious mistake to avoid.
Access comes through profiles like Night Shift or Day Shift, so a change to the shift pattern happens once, not twenty-three times.
Automated rules
The best thing an estate tool can do is stop needing a person. Automations arm each site on a schedule (set at 19:00, unset at 07:00) and run the lighting in the same place.
The record
The log splits into Alarms, Errors and Arm/Disarm, groups by month, and names the person: Set by Noah Johnson, Feb 10th 11:45:32.
Availability
Taking on freelance projects
Replies within 1–2 days
Location
Remote · Bulgaria
Local time
The first call
Free, and commits you to nothing
Elsewhere