Building a unified view of facilities for a renewable energy company
ju:niz builds sustainable real estate and renewable-energy production, storage and conversion. As it scaled to 25 plants and 260 megawatt-hours of capacity, the monitoring systems that had served a handful of sites began to crack — and the data that mattered most was the hardest to reach.
The challenge
Traditional monitoring gave limited resolution: edge databases running InfluxDB v1 captured a reading only every five seconds, and when something went wrong the data was effectively inaccessible without dispatching an engineer to the site. Alerting was coarse, too — among many identical devices, the system couldn't say which specific component had failed.
Ricardo Kissinger, Head of IT Infrastructure and Security, set out to build a centralized, data-driven monitoring system. The migration ran straight into friction: unreliable custom Python scripts for edge-to-cloud transfer, InfluxDB version incompatibilities from v1 to v3, manual edits required to migrate Line Protocol files, and rate limits of roughly 2 MB/s on InfluxDB v3 Cloud clusters that made moving data between buckets painful.
Moving data around is also not easy. Now we collect up to 400 GB of compressed Influx files per day on the biggest plants.
Ricardo Kissinger — Head of IT Infrastructure & Security, ju:nizThe solution
Ricardo chose Quix to integrate data from every source into a single InfluxDB v3 Cloud production bucket. Quix pulled data from the existing InfluxDB v1 and v2 instances and new plant sources via Telegraf, reprocessed data out of the Dev and Test buckets, and centralized all device monitoring into one place — so a single alert rule per device type could cover every project.
Quix Connectors and templates simplified the integration, pre-processing restructured data coming from different InfluxDB versions, and backpressure support gave precise control over ingestion rate so the clusters never overloaded. The added granularity meant ju:niz could finally monitor individual components — right down to single battery cells.
I faced complicated issues until I found Quix. It just works — so much more professional than before. And Quix saves me a lot of time. I'm not a full-time programmer, so I'm using Quix templates and AI.
Ricardo Kissinger — Head of IT Infrastructure & Security, ju:nizThe results
The data infrastructure was transformed. ju:niz went from a 20 GB total database in February 2023 to collecting 1,000 GB a day — a 1,500× increase in capacity — while cutting storage costs by a factor of 100, now processing 30TB a month for what 300 GB used to cost.
Monitoring became granular and device-specific, pinpointing the exact battery cell rather than a vague device failure, and the pipeline scales easily as the company grows. Connectors and templates keep development simple enough for a team that isn't staffed with full-time programmers.
Now, using Quix, we have the status of every module, cell and rack. We know exactly what's going on. We collect data points from all the plants and databases and write them into a single bucket — which lets us create one alarm rule per device type for all projects, significantly reducing our workload.
Ricardo Kissinger — Head of IT Infrastructure & Security, ju:niz