Industrial IoT · Real-time

The control plane for your factory data.

Ingest sensor streams, transform them in real time, alert on anomalies, and reason about production with an AI agent — all from one cloud platform.

NATS · JetStream · ClickHouse · Kubernetes-native
ingest.sh bash
 # Stream a sensor reading into indu.cloud
nats pub sensors.line-3.temperature \
  '{"ts":1747713600,"value":74.2,"unit":"°C"}'
line-3 · temperature
live
min 68.4
avg 72.1
max 76.8
Trusted by manufacturing teams
ACMENORDFABKILNCOAXLE-9POLYMER+STAHL
Product

One platform, from sensor to decision.

indu.cloud brings together every layer of the modern industrial data stack — ingestion, transformation, monitoring, and AI — without the integration tax.

Real-time monitoring

Sub-second sensor dashboards over WebSocket. Pin the views your operators need.

Alerting that fires

Rule-based engine with email and Slack. Catch threshold breaches and anomalies the moment they happen.

Stream processing

Twelve built-in transforms — deadband, EMA, rate-of-change — applied at the edge of your data.

AI agent + predictive

Ask questions about production in plain English. Forecast failures before they take a line down.

Real-time monitoring

Watch every signal as it happens.

  • Sub-second WebSocket updates — no polling, no stale dashboards.
  • Compose views from any device, any sensor, any tag.
  • Pin favorites for operators; lock layouts for the line.
Screenshot · 01-dashboard
Real-time historian dashboard
Alerting

Catch the breach, not the blame.

  • Threshold, equality, contains and custom expression rules.
  • Multi-channel delivery: email via SES, Slack webhooks.
  • Live alert feed with leasing and ack workflows.
Screenshot · 02-alerts
Alert rules + live event stream
Stream processing

Shape the data before it lands.

  • Deadband, moving average, EMA, rate-of-change, rolling std-dev.
  • Stateful and horizontally scalable — survives a node going down.
  • Outputs persist to ClickHouse for historical queries.
Screenshot · 03-streams
Stream transformation editor
AI agent + predictive

Ask production a question.

  • Chat with an agent that knows your devices, sensors and history.
  • Predictive maintenance models surface failures before they happen.
  • Anomaly detection runs in the background — no manual thresholds.
Screenshot · 04-ai-chat
AI agent answering a production question
How it works

A clean path from machine to insight.

Built on NATS, ClickHouse and Kubernetes — boring infrastructure, so the interesting parts can move fast.

Architecture diagram
Devices → NATS → Streams → ClickHouse → Dashboards / Alerts / AI
01
Ingest

Devices publish over NATS or AWS IoT. No agents to babysit.

02
Transform

Streams pass through configurable filters and aggregations in flight.

03
Store

Everything lands in ClickHouse — query years of history in milliseconds.

04
Act

Dashboards, alerts, predictive models and the AI agent all read from the same source of truth.

Built for engineers

Talk to your machines from anywhere.

Push from a PLC bridge, an OPC-UA gateway, AWS IoT, or a one-line NATS command. Pull through REST or WebSocket. No proprietary SDK required.

  • · NATS JetStream native
  • · AWS IoT bridge
  • · REST + WebSocket query API
  • · OpenTelemetry traces & logs
publish.sh bash
 # Publish a sensor reading
nats pub sensors.{device}.{sensor} \
  '{"ts":1747713600,"value":74.2}'
query.http http
GET /v1/sensors/line-3/temperature?from=-1h&bucket=10s
Authorization: Bearer $INDU_TOKEN 
stream.yaml yaml
transform: moving_average
window: 30s
input: sensors.line-3.temperature
output: derived.line-3.temperature.ma30

See your factory, live.

Book a 30-minute walkthrough. We'll connect a simulated line and show you what indu.cloud looks like with your data.