No-code Industrial AI for process teams.
One workspace for engineers and operators to explore plant data, build models, monitor live process behavior, and ship recommendations to the control room — without a single line of Python.
- 01liveData Explorer38 tags streaming
- 02okNo-code model builder12 models in workspace
- 03liveAnomaly detectionanomaly-v3.1 · CMP_042
- 04liveForecastingyield-forecast-v2
- 05watchPrescriptive optimizationsetpoint-opt-v1
- 06okClosed-loop automationguardrails active
One workflow, from raw sensor data to closed-loop control.
- 01→
Connect data
Historians, SCADA, OPC UA, MQTT — secured by the Intelecy Gateway.
- 02→
Explore / process context
Browse tags, overlay variables, find the signal that matters.
- 03→
Build models
Drag-and-drop templates. Engineers own the model end-to-end.
- 04→
Deploy & monitor
Live anomaly detection and forecasts running 24/7 with drift alerts.
- 05
Recommend / automate
Operator-facing setpoint guidance, with optional write-back to the DCS.
Six modules, one workspace.
Every module shares the same data, the same tags, and the same governance — so a signal you spot in Data Explorer becomes a model, an alert, and a setpoint recommendation without leaving the page.
Data Explorer
High-speed trend analysis across thousands of tags, with multivariable exploration, correlations, and side-by-side historical vs live data.
- Browse and overlay any tag from Historian or live stream
- Pearson / lag correlations across signals
- Side-by-side historical and live windows
- CMP_042.vib
- DRY_B.temp
- LINE_02.flow
- BLR_01.kWh
- TANK_3.lvl
No-code model builder
Templates for anomaly detection, forecasting, and prescriptive recommendations. Engineers train, validate, and monitor models without leaving the workspace.
- Templates for the three model families plant teams need
- One-click training with built-in validation
- Drift monitoring out of the box
Anomaly detection
Detect deviations before they turn into downtime, quality issues, or safety incidents — based on the normal operating envelope your engineers approve.
- Multivariate residual models, not just thresholds
- Operator-friendly explanations per alert
- Quiet hours, severity tiers, escalation rules
Forecasting
Predict future process values and outcomes in real time — yield, energy demand, effluent loads, batch end-states — so operators act, not just observe.
- Horizon configurable from minutes to hours
- Confidence bands tuned per signal
- Backtested against historical periods
Prescriptive optimization
Recommend the next setpoint move, with confidence and a plain-language explanation of why — so operators can decide in seconds, not shifts.
- Recommended action with expected impact
- Confidence score and contributing variables
- Suppressible per operator, per shift
- • Inlet humidity trending below 7-day average
- • Residual on dryer outlet temp +1.8σ
- • Forecast shows yield stable in proposed range
Closed-loop automation
Promote recommendations to automated setpoint updates when you're ready — with human-in-the-loop approval, hard guardrails, and a full audit trail.
- Operator approval or fully automated, per loop
- Hard min/max guardrails enforced before write-back
- Every change logged and replayable
- 14:22 · approved · −1.4°C · m.lien
- 13:55 · auto · −0.6°C · within guardrails
- 13:10 · dismissed · low confidence (74%)
Fits the plants and protocols you already run.
The Intelecy Gateway connects securely to your existing Historian, DCS, and SCADA — without ripping out anything that already works. Deploy on-prem, in our EU-hosted cloud, or hybrid; you decide where data and models live.
Built to clear your enterprise security review.
Your data and models stay yours. No training on customer data, ever.
Role-based access per workspace, per model, per tag — managed by your IT.
At-rest and in-transit. Customer-managed keys available on enterprise plans.
Every model change and setpoint write logged. Export models any time.
- Data ownershipcustomer-owned
- EncryptionTLS 1.3 · AES-256 at rest
- Tenant isolationper-workspace
- Access controlrole-based, per tag
- Model trainingnever on customer data
- Audit logevery change replayable
How data is protected and who owns it: encryption, per-workspace isolation, tag-level access control, and an auditable change history.
From kickoff to control-room in one quarter.
- Day 1–2→
Connect
Intelecy Gateway installed. Historian tags streaming into the workspace.
- Week 1→
First model
Anomaly or forecast model built with your engineers, validated against history.
- Weeks 4–6→
Value hypothesis
Quantified savings vs baseline — downtime, kWh/ton, scrap, throughput.
- Quarter 1
Adoption
Recommendations live in the control room; first closed-loop guardrails defined.
Questions buyers ask before the demo.
Who uses Intelecy?+
Process engineers and operators own day-to-day use. Plant leadership tracks outcomes; IT/OT owns the Gateway and access controls.
What data is needed to get started?+
Most customers start with 20–50 tags from an existing Historian (PI, AVEVA, AspenTech) covering one unit. No new sensors required for the first model.
How are models validated?+
Every model is backtested against historical periods you choose, with operator-readable metrics (residual MAE, false-positive rate) before it goes live.
Where do recommendations run?+
Recommendations are served to operators in the Intelecy workspace and can be surfaced in your existing HMI via API. They never bypass the operator unless you enable closed-loop.
How is closed-loop governed?+
You define hard min/max guardrails, max change per interval, and rollback conditions per loop. Every write-back is logged and replayable; operators can pause closed-loop at any time.
Want a walkthrough on your own data?
A 30-minute working session with our process AI team. Bring a unit, a goal, and a few tags — leave with a working model proposal.