Plan, build and operate. One estate model across all of it.

Every block reads and writes the same record of your datacenters. Click a block to see what it holds.

SynapseDC platform

Plan & build sellable capacity → commissioned hall

The core one estate model

Operate impact and cost on every fault

Govern & data
Governed end to endChange request on every write · Approval-gated actions · Audit trail · Every count derived
Deploy where your data must liveOne platform, your choice of footprint, inside your identity stack.
On-premisesYour cloudAir-gappedEmbedded in your portal

One estate model. The graph every block reads and writes: building, plant, racks, devices, ports, cables, GPUs, services, customers and contracts. Every count is derived from it, never typed in.

  • 51 entity types, 107 relationships
  • Stable identities, so adding data never reshuffles the rest
  • Import template in load order for onboarding
Explore

Govern

Prove the estate meets the frameworks you are audited on.

Architecture

One collector per site. One model for the estate.

Read-only southbound, outbound-only from the site, and the same REST contract the screens use for your own tools.

Inside your datacenterread-only · no inbound ports
  • BMS · UPS · PDUs · metersBACnet · Modbus · SNMP
  • CRAH · CDUs · sensorsBACnet · Modbus
  • Routers · switches · opticsSNMPv3 · gNMI · syslog
  • Servers · BMCsRedfish · IPMI
  • GPUs · scheduler · fabricDCGM · Slurm · UFM
  • Doors · cages · CCTVPACS · VMS
On-site collectoractive + standby
  1. Read each source with a read-only role
  2. Normalise alarms; keep source and time
  3. Buffer when the uplink drops
  4. Send outbound only, one channel
Never controls plant · never polls tenant gear
SynapseDCone estate model · one alarm model · one money model
IngestOTLP · remote write · events
Storestime series · events · records
Estate modelsites → racks → devices → ports, customers, contracts
Read models
ImpactRevenueRoot causeGPUPhysicalCapacityPower economicsSecurityCompliance
Opstral AIanswers · causes · actions, approval-gated
Screens · API · eventsthe same REST contract the screens use · CloudEvents
Your systemsthe module that owns a record stays its system of record
  • ITSM · incidents, changes, tickets
  • CRM · customers, contracts, notices
  • IPAM · DNS · DHCP
  • SIEM · OCSF events
  • Your tools · REST API v1 · webhooks · Kafka
Runs where your data must liveOn-premisesYour cloudAir-gappedYour identity · OIDCEmbedded in your portal

Design principles

Rules the whole product keeps.

Counts are derived, never typed in

Every number comes from the entities behind it, so the same figure reads the same on every screen.

An alarm counts only where it belongs

Site, edge, dependency or downstream. Downstream symptoms appear under the incident that caused them.

One issue per impacted service

Each service is charged to its most severe cause, so money reconciles across issues, customers and sites.

Health shown calmly

LEDs, beacons, halos and chips. Red and amber are used only for faults.

Every record change has a change request

Standard pre-approved work, an open CR at the site, or a new one raised for approval.

The platform module owns the record

CRM, ITSM, inventory and projects stay where they are. SynapseDC builds what is datacenter-specific.

Opstral AI on every screenAnswers, likely causes and recommended actions from your own data.Learn more

See your own estate in it.

Bring the fault your team keeps dealing with. We walk it end to end on a demo estate shaped like yours, then agree what a one-site pilot should prove.