Replaced an eight-page energy report with one daily list for the operators of 40 commercial buildings
Client snapshot
An energy services firm that monitors heating, cooling and lighting systems across about 40 commercial buildings for its clients. A team of six building operators responds to what the data shows, supported by a small analytics team.
The problem
The analytics team had built a thorough Power BI report: eight pages of charts on consumption, setpoints, alarms and equipment runtime. It was accurate and nobody used it the way it was meant to be used. Operators are technicians, not analysts. Finding the three buildings that needed a visit meant clicking through every page and comparing charts by eye. Most mornings it took around 40 minutes, and plenty of anomalies, such as a chiller running all weekend, were spotted a week late or not at all.
What we built
A daily list, built from the same data the report used.
Every morning at 6am, n8n pulls the last 24 hours from the building management system and the data the Power BI model already holds. It checks each building against its own normal pattern and its schedule. Claude turns what it finds into a short list of 10 to 15 items in plain language: "Building 12, chiller 2 ran 31 hours over the weekend against a weekday-only schedule. Check the time program." Each item has a priority, a recommended action and a link to the exact page and filter in the Power BI report.
The list lives in Airtable. Operators mark items done, assign them or note what they found, so the analytics team can see which alerts were useful and which were noise. The thresholds were tuned with the operators in the first month.
A weekly view shows which buildings and which kinds of problems keep coming back.
Results, 90 days in
- Finding what needs attention each morning: about 40 minutes → about 5.
- Energy anomalies acted on within a day: around one in three → most of them.
- Operators work from one list instead of eight report pages; the report is still there when they need detail.
- The weekly view surfaced four buildings with recurring schedule faults, fixed once instead of chased every week.
Quote
"I don't need to read the charts any more. The list tells me which building and what to check when I get there." Building operator
Build snapshot
- Timeline: 14 business days
- Stack: building management system API, Power BI, n8n, Claude, Airtable
- Investment: $5,500
- Who owned it: the operations lead owns the thresholds; the analytics team reviews the feedback on alerts monthly.
The pattern
This is the pattern we ship in the Operator's Dashboard engagement: read across the data, and put the things that need a person on one list. If your team owns a report nobody reads, the build is the same, with your sources swapped in.
This is an illustrative composite drawn from operator engagements. Names, identifying details, and exact metrics are anonymised. The workflow, stack, and approach are real.
Looks like your situation?
A 30-minute call tells us whether the same pattern fits your operation, and what it would cost to ship it.