↰ work

HClimate Dashboard

Turning portfolio-wide BMS telemetry into a working diagnostic surface — from point-level noise to the handful of assets a facilities manager should touch today.

HClimate — building climate dashboard
Project

HClimate monitors building climate across an instrumented property portfolio — temperature, humidity and device state, sampled per point across every site on the estate. We designed and built the diagnostic layer on top of that telemetry: a portfolio triage view and a device-level drill-down that let a facilities team find the asset behind a comfort complaint without exporting a spreadsheet.

Our role

Product Design, Front-end Engineering, Data Visualisation

Client

HClimate

Year

2026

Industry

PropTech / IoT
React (Vite) SPA

Team

Design & build: Mario Ushev
created at Paragon Studio

A high point count is not the same as visibility

An instrumented portfolio produces telemetry continuously — temperature, humidity and device state, sampled per point across every building on the estate. What an energy or facilities manager is actually measured on is narrower than that stream: comfort complaints and hot/cold calls, energy intensity per m², and how long a fault runs before anyone notices it. None of those are answered by a raw point stream.

Point count is the easy part to sell and the hard part to use.

why instrumentation stalls

In practice the workflow degrades to CSV export and a pivot table, and detection becomes reactive — the tenant calls before the dashboard says anything. That is the failure mode worth designing against: plant running out of schedule for days, a drifting setpoint, or a dropped sensor quietly biasing an average nobody re-checks.

Underneath sat a commercial constraint. A platform that stalls when someone filters it is a platform that stops being opened, and adoption is what the whole investment rests on: unused analytics cannot justify the sensor capex at renewal. Speed here is not a nicety — it is the difference between a monitoring contract that renews and one that is written off.

Estate-level triage, device-level proof

Our mission was to build the layer that makes the instrumentation pay for itself — a tool with scientific credibility and consumer-grade responsiveness, structured the way an estate is actually managed: triage at portfolio level, then drill to the device that explains the reading. It had to stay fast at full estate scale rather than at demo scale, because the moment it hesitates is the moment a facilities manager goes back to the spreadsheet.

Scope
  • Product design
  • Portfolio overview
  • Device-level drill-down
  • Fault investigation workflow
  • Time-series visualisation
  • Performance at estate scale

Two views that mirror how an estate is actually walked

The app resolves to two views because that is the shape of the job. A Buildings overview answers "which sites are behaving"; an All-devices breakdown answers "which unit is causing it" — triage, then evidence, with no dead ends between them.

The decision that mattered was refusing to let estate size show through to the user. However many buildings and devices sit behind a view, the interface responds at the same speed — so an estate can be onboarded without the tool degrading, and a facilities manager never learns to expect a wait. Interrogating the portfolio stays a reflex rather than a chore, which is the only condition under which anyone checks it daily rather than after a complaint.

A tool that stays open during the working day

Interrogating the estate is now a first-class action rather than an export job: a facilities team moves from portfolio to device without waiting on the interface, which is the precondition for shortening time-to-detect on anything at all.

2
views: portfolio triage → device evidence
0
CSV exports in the fault investigation path

Built with

React Vite Web Workers JavaScript CSS

Explore more work

view all →

Your project
could be next

We design and build websites and products for ambitious brands. Let's make it happen.

let's talk