Signals
Decision intelligence that detects patterns, surfaces insights, and guides teams toward smarter actions.
At a glance
Real-time anomaly detection
Real-time anomaly detection on governed KPIs with context and root-cause hints.
Forecasts and scenarios
Forecasts and scenario simulation for short and medium horizons.
Next best actions
Next best actions delivered as clear playbooks with owners and due dates.
Smart subscriptions
Subscriptions and alerts to email, Slack, Teams, and webhooks.
Native integration
Native integration with Fabric for metrics and with Guard for policies.
Why Signals
Dashboards show what happened after the fact. Enterprises need a system that notices change as it occurs, predicts what comes next, and recommends what to do. Signals converts trusted data into timely guidance so teams move faster with confidence.
What Signals does
Define
Create governed KPIs from Fabric's semantic layer with proper contracts and ownership.
- •Create governed KPIs from Fabric's semantic layer.
- •Metric contracts with owners, thresholds, and business calendars.
- •Dimension guards to respect access policies from Guard.
Detect
Advanced anomaly detection with automated narratives and context explanations.
- •Streaming and batch anomaly detection with seasonality awareness.
- •Change-point, trend shift, and outlier identification with confidence scores.
- •Automated narratives that explain what moved and where.
Forecast
Predictive modeling with scenario simulation and accuracy tracking.
- •Short-term and medium-term forecasts with backtesting.
- •Scenario inputs for price, volume, and mix with sensitivity analysis.
- •Reliability bands and accuracy metrics for each model.
Recommend
Intelligent playbooks that map patterns to actionable next steps.
- •Playbooks that map patterns to next best actions.
- •Assign owners and SLAs, capture approvals where required.
- •Optionally trigger workflows in Studio for automation.
Act and learn
Execute actions and learn from outcomes to improve future recommendations.
- •Send tasks to Slack, ticketing, CRM, or email with one click.
- •Log outcomes, measure impact, and refine playbooks over time.
- •Close the loop with feedback to improve detection and routing.
AI inside Signals
Pattern intelligence
Time-series models detect seasonality and regime shifts.
Root-cause hints
LLM summaries review movements across segments and propose likely drivers with citations.
Action ranking
Policies and historical outcomes inform the ranking of recommendations.
Continuous learning
Feedback improves thresholds, segmentation, and model choice.
Architecture overview
Comprehensive decision intelligence architecture from metrics to actions with full observability and security.
Signals turns governed KPIs into timely guidance through a pipeline of define, detect, forecast, recommend, and act with learning. Guard enforces policy end to end and ModelOps ensures quality and cost control.
Common use cases
Revenue operations
Pipeline drop in a region, attribution shifts, win-rate drift, quota pacing.
Finance
Expense spikes, variance to budget, cash-collection risk, late invoice forecasts.
Supply and inventory
Stockouts, demand surges, lead-time changes, vendor performance.
Risk and fraud
Unusual transaction patterns, velocity breaches, device reputation shifts.
People analytics
Offer acceptance trends, attrition risk, overtime anomalies.
Customer support
Ticket backlog surge, sentiment change, escalations forecast.
Example playbooks
Revenue dip in APAC
Verify data freshness, compare channel mix, increase remarketing budget, and alert regional lead.
Late invoice risk
Notify account owner, schedule reminder to customer, and forecast cash impact.
Inventory surge
Freeze nonessential POs, rebalance stock, and notify fulfillment.
Churn risk rise
Trigger Studio to draft outreach, open a ticket, and add a save-offer task.
Data quality regression
Roll back the latest pipeline version and notify Fabric owners.
How it works, end to end
Select KPIs from Fabric and define contracts, thresholds, and owners.
Signals profiles history and enables detection with seasonality.
Forecasts and scenarios are generated with accuracy tracking.
When a pattern occurs, Signals sends an alert and proposes next actions.
Owners accept, edit, or route to Studio to automate tasks.
Outcomes are logged and playbooks are refined for the next cycle.