Accentrust

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.

Decision intelligence visual placeholder

What Signals does

1

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.
2

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.
3

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.
4

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.
5

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.

Inputs
Fabric Metrics & Dimensions
Governed metrics, dimensions, contracts
Warehouses & Lakes
Snowflake, BigQuery, Redshift
External Events
Product telemetry, transaction logs, 3rd party
Business Calendar
Workdays, holidays, earnings cycles
Define (Definition & Contracts)
KPI Catalog
Metrics inventory & classification
Metric Contracts
Definitions & thresholds
Owners & Calendars
Owners & SLAs
Dimension Guards
Access control
Detect (Detection)
AI
Anomaly Engine
Seasonal & trend detection
Change Monitor
Version & config changes
AI
Root-cause Hints
Dimension breakdown & correlation
Signal Store
Standardized signal events
Forecast (Forecasting & Scenarios)
AI
Forecast Library
ETS,SARIMAX,GBM
Scenario Simulator
Price & volume sensitivity
Accuracy Tracker
Backtest error assessment
Forecast Store
Forecast trajectories & intervals
Recommend (Recommendations & Playbooks)
Playbook Rules
Configurable rules
AI
Policy-aware Ranker
Historical effectiveness ranking
AI
Recommendation Engine
Actionable recommendations
Approvals
Human review & approval
Act & Learn (Execution & Learning)
Action Hub
Slack•Teams•Jira
Studio Triggers
Workflow triggers
Outcome Logger
Record business outcomes
AI
Feedback Loop
Closed-loop learning
Channels
Slack / Teams
Jira / ServiceNow
Salesforce / HubSpot
Email / SMS
BI & Dashboards
Studio Workflows
Safety & Governance (Guard)
Access Control
Row & column level access
Privacy & Redaction
PII processing
Policy Enforcement
Recommendation validation
Audit Logs
End-to-end logging
Observability & ModelOps
Traces & Costs
Throughput, latency & cost
Backtests
Forecast backtest evaluation
Quality Gates
False positive/negative thresholds
Drift Monitor
Data & concept drift
Experiment Manager
A/B testing & experiments

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

1

Select KPIs from Fabric and define contracts, thresholds, and owners.

2

Signals profiles history and enables detection with seasonality.

3

Forecasts and scenarios are generated with accuracy tracking.

4

When a pattern occurs, Signals sends an alert and proposes next actions.

5

Owners accept, edit, or route to Studio to automate tasks.

6

Outcomes are logged and playbooks are refined for the next cycle.

Ready to turn data into decisions?

Transform patterns into action with decision intelligence that detects, predicts, and guides your teams.