AI Ops Console
A conversational control plane for industrial operations. Operators speak to the runtime in natural language; the console plans, dispatches, and reports across the agent fleet — with KPI grounding and policy gating.
Dispatch an autonomous workflow
Authenticate with a runtime key, declare the workflow, stream execution events as they unfold across the agent fleet.
from wavelet import Runtime
rt = Runtime(api_key="wlt_live_...")
execution = rt.workflows.run(
workflow="predictive_maintenance",
asset="LINE-07-COMPRESSOR-A",
context={"severity": "high", "auto_dispatch": True},
)
for event in execution.stream():
print(event.phase, event.agent, event.payload)Operational Workflow Templates
precision_manufacturingPredictive Maintenance
Enhance, Not Replace
Bidirectional runtime sync with the systems your operation already runs on — SAP, Oracle, MS Dynamics, MES, SCADA, EAM, IoT.
Runtime Evolution
Hallucination detector accuracy +6.3 pp; new auto-patch policy for ERP write failures.
Multi-agent deliberation now supports cross-domain risk arbitration.
Vector index throughput +2.4×; new industrial ontology pack (energy, semiconductor).
Execution scheduler rebuilt — p99 latency 412ms → 184ms.