Exploratory directions. No dates. No promises.
Model-centric CDE
Today's CDEs are document stores with a viewer bolted on. We're building the inverse — the model is the document. Drawings, specs, submittals, RFIs, photos, and maintenance records hang off elements, not the other way around. Instead of searching for the right PDF, you click the wall.
Imagine opening a room and seeing its 2023 fire inspection, the HVAC submittal PDF, the supplier's spec sheet, and last month's maintenance photo — all already attached to the space, no file hunt required.
Connectors Where Work Already Happens
SharePoint, Microsoft Teams, Outlook, Office 365. A model edit posts into a Teams channel; a SharePoint document binds to an IFC room; an Outlook thread attaches itself to the component it references. Not yet-another-place to log in — a surface that meets each team in their existing tool.
Imagine resolving a BCF topic in-browser and the thread appears as a Teams message in the project channel, with the viewpoint image attached and the component already linked.
Real Digital Twin, Not the Brochure Version
Live sensor telemetry wired to the IFC spatial structure. Historian queries, anomaly detection, failure prediction — scoped per room, system, and component. A building that can answer what its own state is right now, and what broke last Tuesday at 3am.
Imagine clicking an HVAC unit and seeing its last 90 days of runtime hours, current inlet temperature, and the three anomalies flagged since March — all pulled from a live sensor feed bound to that exact element.
Sensor-Driven Maintenance
Maintenance schedules that adapt to actual use instead of a paper calendar. A pump's vibration pattern triggers a work order before it fails; a filter's pressure drop books its own replacement; a chiller's runtime meter decides when it's due. The IFC model is ground truth, sensor data is the heartbeat, the schedule is the output.
Imagine a pump accumulating 1,000 hours of runtime: the platform creates the service ticket, attaches the manufacturer's procedure, and picks an inspection window from your facility team's free slots — before the bearing complaint arrives.
Agentic AI Automations
Agents that do work, not just answer. Preventive maintenance tracked per component — the next due date surfaces on the element, and the platform drafts the inquiry email to the responsible maintainer. A sensor anomaly writes the incident document, tags the affected rooms, and alerts the owner. Every outbound action requires explicit human approval — agency with a leash.
Imagine a chilled-beam flagging a temperature-delta anomaly overnight: by morning, the platform has drafted an incident report, attached the sensor trace, linked the affected rooms, and queued an email to the operator — all waiting for one click of approval.
Knowledge Platform
Every RFI, photo, comment, BCF topic, supplier spec, and sensor reading indexed by the model. Ask what you know about a room's HVAC and get a single answer pulled from years of project history. Institutional memory that happens to be geometry-aware — so new team members don't need to ask the old hands where anything is.
Imagine a new facility manager on day one asking "what issues has Room 3.14 had?" — one answer, citing the 2024 leak RFI, the 2025 BCF topic, and the latest sensor trend, all with links back to the element in the model.
Open Ecosystem
API-first and webhook-native. Every action the UI can perform is a documented endpoint; every state change can fire a webhook. Plug in custom workflows, third-party analytics, or your own in-house tools — the platform is a backbone, not a walled garden.
Imagine a webhook that POSTs to your ERP whenever a maintenance ticket closes, and a custom agent that queries the model API to auto-generate the monthly facility report — both running without waiting for us to build them.