Saltar al contenido principal

Analytics Dashboard

TietAI's Analytics Dashboard provides real-time visibility into your organization's clinical operations, pipeline performance, and patient outcomes. It includes configurable modules, an AI-powered copilot for ad-hoc queries, and streaming data from pipeline executions.


Dashboard overview

The Analytics Dashboard is accessible from the main navigation under Analytics. It presents data in configurable workspaces — each workspace contains one or more modules (visualizations, tables, or metrics).

Default modules

ModuleWhat it shows
Pipeline executionsTotal runs, success rate, failure rate, average duration — filterable by date range
Patient sync statusNumber of patients synced per EHR connector, last sync time, error count
Agent activityAI agent invocations, tokens consumed, average response time
Wearable data volumeReadings ingested per platform, per day
Audit summaryLogin events, data access events, admin actions — last 24 hours

Dashboard Copilot

The Dashboard Copilot is an AI assistant that helps you explore your data using natural language. Open it by clicking the Copilot button in the Analytics toolbar.

What the Copilot can do:

  • Answer data questions — Ask "How many pipelines failed last week?" or "Which connector has the highest error rate?" and get instant answers with supporting data
  • Add dashboard modules — Ask "Show me a chart of daily patient syncs for the last 30 days" and the Copilot creates a visualization module on your dashboard
  • Run ad-hoc queries — The Copilot can execute SQL queries against your analytics data and return results as tables or charts
  • Export data — Request data extracts in CSV, JSON, or Parquet format with configurable date ranges

Copilot actions

The Copilot responds with structured actions that modify your dashboard:

ActionWhat it does
add_moduleAdds a new visualization module to the current workspace
update_moduleModifies an existing module's query, filters, or visualization type
remove_moduleRemoves a module from the workspace
create_workspaceCreates a new workspace with specified modules
run_queryExecutes an ad-hoc SQL query and displays results
extract_dataGenerates a downloadable data extract (CSV, JSON, or Parquet)

Real-time streaming

The Analytics Dashboard supports real-time data streaming for live operational monitoring:

  • Pipeline events — See pipeline execution events as they happen via NATS event bus integration
  • WebSocket updates — Dashboard modules update in real time without page refresh
  • Live agent activity — Monitor AI agent conversations and inference requests as they stream in

Enable real-time mode by clicking the Live toggle in the dashboard toolbar. Modules with streaming support show a green pulse indicator.


Workspaces

Workspaces let you organize dashboard modules by topic or team.

Creating a workspace

  1. Go to Analytics → Workspaces → New Workspace
  2. Name the workspace (e.g., "Clinical Operations", "IT Monitoring", "Executive Summary")
  3. Add modules by clicking Add Module or asking the Copilot
  4. Arrange modules by dragging them into position
  5. Click Save

Sharing workspaces

Workspaces can be shared with other users in your organization:

  • Can view — See the workspace and its data
  • Can edit — Modify modules and layout

Data sources

The Analytics Dashboard aggregates data from across the TietAI platform:

SourceData available
Pipeline executionsRun count, duration, status, per-node metrics
FHIR storePatient count, resource counts by type, storage usage
ConnectorsMessage throughput, error rates, latency
AI agentsInvocation count, token usage, response times
WearablesReading count by platform and data type
Audit logsUser activity, API usage, security events
AuthenticationLogin counts, failed attempts, active sessions

Custom reports

Beyond the live dashboard, you can generate scheduled reports from analytics data:

  1. Go to Analytics → Reports → New Report
  2. Select a template or start from scratch
  3. Configure data sources, filters, and date ranges
  4. Choose output format: PDF, CSV, or email summary
  5. Set a schedule (daily, weekly, monthly) or generate on demand
  6. Click Save

Reports are delivered to the configured recipients via email or stored in the Reports section for download.