Real-time dashboards and data visualisation

Dashboards that turn streaming operational data into decisions: live KPIs, alerts and drill-downs, designed around what the operator needs to act on next.

Most dashboards fail because they try to show everything at once, not because the data is wrong. I start from the opposite end: what is the one number this person has to act on in the next five minutes, and what can be left out.

Technically that means streaming updates instead of polling, a query layer sized to what each view renders, and honest handling of bad connections. A dashboard that silently shows stale data is worse than one that shows nothing, so connection health and stale-data detection are part of the design, not an afterthought.

I build these in React, with Grafana where domain experts need full control over time-series queries and custom components where the workflow matters more than the chart.

What this covers

  • Streaming data over WebSockets and realtime subscriptions
  • GraphQL gateways over legacy and third-party systems
  • Alerting that highlights only actionable deviations
  • Connection health, stale-data detection and recovery
  • Grafana integration and custom charts with ECharts
  • Access control per user and per metric

Where I did this

  • Nautical Dashboard (2025): Fleet management dashboard that turns AIS, fuel, engine and telemetry data into one operational view, with any vessel two clicks away.
  • Realtime Dashboard (2025): Modular analytics dashboard with streaming KPIs, first-class alerts and drill-downs, updating tiles in under 200 ms.
  • GeodataPro (2025): Health data portal with fast, filterable dashboards built for professionals on mobile devices in the field.
  • Digital Signage Studio (2025): KPI, capacity and custom ECharts widgets that keep live data on screens without anyone editing the page again.

By Guus van de Wal, Senior Front-end Developer & UI Engineer.