One license. Every capability.

Relay

Relay connects legacy databases, HL7 feeds and scanned documents and turns them into validated FHIR resources, without a data migration.

• Bidirectional Sync

  Maintain real-time consistency between legacy HIS and your modern FHIR data layer.
• AI Document Intelligence

  Extract structured clinical data from scanned PDFs, lab reports, and handwritten notes.
• Zero-Code Mapping

  Visually map legacy database fields to FHIR resources using our AI-guided interface.
• Schema Discovery

  Automatically detect legacy database structures and suggest the best FHIR mappings.
• Error Replay

  Dead-letter queues for failed messages with one-click fix and replay capabilities.
• Data Lineage

  Trace every FHIR resource back to the exact HL7 segment or PDF page it originated from.

Ingestion Protocols

HL7 v2.x, MLLP, DICOM, SFTP

Mapping Engine

AI-Assisted Semantic Mapping

Engine

Fhirlace, deterministic runtime

Output Format

FHIR JSON / XML

Relay runs on the Fhirlace engine. AI drafts each source-to-FHIR mapping at design time and a person approves it; at runtime the mapping is deterministic, so the same input always produces the same resource. Scanned documents go through OCR and clinical extraction, and every resource keeps a pointer back to the segment or page it came from before it is stored in Fabric.

Sage

Sage indexes the patient record, your protocols, SOPs and the content you license, and answers questions with the exact passages it used. If nothing supports an answer, it says so.

• Grounded Clinical AI

  Every sentence in an answer must be supported by a retrieved passage, or it is removed. If nothing supports an answer, Sage says so.
• Natural Language Search

  Ask complex questions like 'How many patients had troponin spikes after surgery?'
• Protocol Indexing

  Upload your hospital's SOPs and clinical guidelines for instant staff reference.
• Source Attributions

  Every answer comes with clickable links to the exact source document or resource.
• Privacy First

  Runs on-premise or in your private cloud with open-weight models you choose. Patient documents stay inside your network.
• Trend Analysis

  Identify emerging clinical patterns across thousands of patient discharge summaries.

Embedding Model

Med-Specialized Vectors

Search Type

Hybrid (Semantic + Keyword)

Context Window

Up to 128k Tokens

Privacy

Air-gapped LLM support

Sage runs a five-stage retrieval pipeline: structure-aware ingestion, hybrid search (semantic plus exact keyword, because drug names and lab codes must match exactly), reranking, evidence verification, and a cited answer. Conductor agents use the same pipeline to gather evidence, so an agent's draft and a clinician's question are grounded the same way.

Pulse

Pulse turns the live record into dashboards, bed and revenue forecasts, and anomaly alerts for hospital leadership. An alert can start a Conductor workflow.

• Real-time Command Center

  Live visibility into every department's performance and patient flow.
• Predictive Bed Mgmt

  Forecast bed occupancy up to 72 hours ahead to plan admissions and staffing.
• Revenue Cycle Intel

  Instantly identify billing bottlenecks and insurance denial patterns to improve cash flow.
• Anomaly Detection

  AI automatically flags unusual clinical patterns or suspicious billing activities.
• Custom Report Builder

  Create and schedule executive reports for your board with one click.
• Multi-facility View

  Aggregate data across your entire hospital network into a single global view.

Data Source

Fabric event stream

Visualization

High-density SVG Dashboards

Predictive Models

LSTM / Prophet for Occupancy

Alerting

Multi-channel (Slack, WA, In-app)

Pulse reads the Fabric event stream into time-series aggregates kept apart from the transactional store, so dashboards and forecasts never slow down clinical systems. Anomalies it flags can start a Conductor workflow, which closes the loop from seeing a problem to acting on it.