Real-time. AI-native. Built for the long haul.
The architectural decisions that let us deliver seven pillars on one platform — and let your data model travel with you as the platform evolves.
Most operations software is a relational schema with screens. Lucendus is an event stream, a semantic asset graph, an AI host, and a workflow engine. With forms and reports dynamically generated from extensible definitions that you or an AI agent can author. Every state change is a business event. Every record carries a Provenance envelope. Every AI output cites the records it read.
Device → Events Host → Stream Bus → Asset Context → AI Host → Insights and Actions
That's the backbone of our system. Telemetry and business data share infrastructure. Dashboards aren't running against a nightly extract — they're rendering against the live event stream and the semantic graph the work runs on. The practical consequence: when a vibration sensor on a critical pump exceeds threshold, the alert fires in seconds, the work order is proposed with the asset's history and last readings already in context, and the audit trail records every step end to end.
AI as a native service, not a feature
AI hosts in Lucendus are first-class workflow participants. They can read the asset graph, propose actions, and (with approval) write back. We don't just ship a "chat panel"; we ship structural AI that does context aggregation, insight generation, predictive recommendations, root-cause analysis, and workflow automation suggestions, with the same provenance and audit story as every human action.
Two MCP surfaces matter: Lucendus is an MCP client (our AI host reaches into third-party systems) and an MCP server (your ChatGPT / Claude / Copilot can reach into Lucendus). That's rare in the operations space and an active part of how the platform lands as the operations layer in agentic workflows.
A unified semantic asset graph
Every entity: assets, signals, work orders, people, locations, vendors, knowledge articles, lives in a single graph. Assets carry multi-dimensional classification: an asset can simultaneously belong to a location hierarchy, a functional hierarchy, an organizational hierarchy, a regulatory scheme (e.g., ISO 14224), and any custom dimensions you define. None of the legacy CMMS and EAM platforms do this; most pick one taxonomy and rebuild the rest as custom fields.
The graph is also where Custom Entity Types live. Customers define their own asset, ticket, person, organizational-unit, and location types — with typed fields, fine-grained validation rules, and bound display/edit forms. Configuration over code; no partner shop required.
The benefits of a modern foundation
Adding a new asset type
Legacy platformsSchema change request, partner ticket, weeks of waiting.
LucendusConfigure a Custom Entity Type in the Lucendus admin pages with fine-grained validation and it's available immediately.
Multi-dimensional asset classification
Legacy platformsPick one hierarchy, rebuild the rest as custom fields.
LucendusNative. An asset is in many hierarchies at once. Create your own custom hierarchies.
"Why did this fail?"
Legacy platformsPull data from CMMS, BI tools, and historian; reconcile in Excel.
LucendusAs the AI host. It reads the graph and the event stream and shows its work along the way.
Reporting period close-out
Legacy platformsWait for nightly extract; reconcile mismatches.
LucendusReports are live and immediately available against the shared data graph.
Custom workflows
Legacy platformsEngagement with an implementation partner. Time and expense.
LucendusAuthor your own workflows in the admin pages. Workflows are version controlled and easy to author and dry-run.
Esposing data to AI tools
Legacy platformsCustom integration or "AI marketplace" add-ons.
LucendusAn MCP server, configured per Lucendus customer. Your AI tools see Lucendus natively and data is protected by fine-grained access controls.
A modern, fast experience. Web and mobile, with fast responses across the platform. Live updates push to every screen without polling.
Real-time everywhere. Assets, signals, work, and analytics all running off a live event stream. Alerts fire in seconds, not after an overnight extract. Reports are always current.
A model that adapts to your operation (extend the data model in the admin without engineering assistance. Define your own asset, work-item, and people types with your own fields and validation. The schema grows as your operation does) unlimited customization without a partner shop.
Sensors and protocols you already have. Connect SCADA, OPC UA, Modbus, BACnet, MQTT, IoT gateways, ERP, productivity, and accounting. Streaming OT/IT bridge is core platform, not an add-on. You bring the hardware; we bring the model.
AI woven through. Agents read across the whole platform with citations and an audit trail. Your existing AI tools (ChatGPT, Claude, Copilot) can talk to Lucendus through MCP with tenant-scoped permission, so the AI workflows you're already building can include your operations data.
Powerful for field staff. Offline-capable mobile experience with photo-native execution, signatures, and the same data model the desktop uses. No "mobile sync" delay; field updates are visible immediately.
Globally accessible, your data close to you. Hosted enterprise SaaS with regional data residency in the EU and US, a global content network for fast access from anywhere, and tenant-branded subdomains for your team. SSO with your existing identity provider.
Backed up, encrypted, audit-ready (automatic backups with point-in-time recovery, multi-region failover, encrypted at rest and in transit. Field-level audit trail across every record. Full provenance on every entry) every record traces back to its source.
Always up to date, no surprise costs. Continuous deployment, no upgrade events, no version sprawl, no maintenance windows. Cost-efficient infrastructure shared across customers; you pay for value (managed assets), not seats.
Want a deeper architectural conversation?
Architecture deep-dive. Schemas, integration patterns, MCP, the workflow engine, the agent runtime. No marketing sales cycle.
- Lifetime early-access pricing — locked for the duration of your relationship
- Direct roadmap influence — your problems drive prioritisation
- Platform configured against your actual data — not a demo environment
- Initial cohort limited to five organisations
Personal review of every application. Response within five business days.