Loading workspace...
Auroranexis documentation
Predictive Intelligence combines operational data — reports, incidents, risks, SLA performance, monitoring, and automation history — into health scores, trend analysis, and forward-looking signals across your client portfolio. Open Dashboard → Predictive to review portfolio summaries and drill into per-client forecasts. Access is gated by the ai_predictive_intelligence plan feature, available on Professional and above.
Predictive Intelligence gives agency leaders a portfolio-level view of client health before problems surface in QBRs or escalation calls. Open Dashboard → Predictive to see overall confidence, executive overview narrative, forecast counts, and key metric cards for churn risk, SLA breaches, incident likelihood, and revenue trajectory.
Each client with sufficient operational history receives a health score, trend direction (improving, stable, declining, critical), and contributing factors drawn from recent activity. Portfolio views highlight clients flagged for attention so account owners can prioritize outreach.
Per-client predictive pages show historical windows (7 days, 30 days, 90 days, 12 months), trend comparisons, incident and risk forecasts, SLA breach probability, and actionable recommendations linked to relevant records. The module checks ai_predictive_intelligence at the plan level — API access via predictive.read scope requires the same feature.
Reactive account management — learning about churn risk only when a client complains — is expensive and damages trust. Predictive Intelligence exists to shift portfolio reviews from backward-looking reporting to forward-looking intervention.
Health scores and early warning signals help operations leaders allocate attention across dozens or hundreds of clients. Rather than reviewing every account equally, teams can focus on declining trajectories, elevated churn probability, or SLA breach forecasts while maintaining lighter touch on stable and growing accounts.
Portfolio signals consolidate data that already lives in Clients, Reports, Risks, Incidents, Monitoring, and SLA modules. Predictive does not replace account owner judgment — it surfaces patterns that are easy to miss when reviewing clients individually.
Contact support@auroranexis.com for onboarding support, billing questions, or product guidance. Include your workspace name, the module you are working in, and a brief description of your goal so we can respond efficiently.
Before a quarterly business review, the account manager opens the client's predictive detail page, notes 90-day trend comparisons, and documents improving report cadence alongside one elevated SLA breach forecast. The QBR deck includes both achievements and proactive items the agency is addressing, backed by health score history from Dashboard → Predictive.
An automation agency newly upgraded to Professional enables ai_predictive_intelligence and opens Dashboard → Predictive. Initial confidence is Low because historical data is sparse. Over six weeks, consistent report publishing and incident recording raise confidence labels. The portfolio owner establishes a baseline health score review as part of the monthly operations cadence.
Every Monday, the operations lead opens Dashboard → Predictive and focuses on the likely churn segment. Three clients show declining health over 30 days. The lead assigns each to its account owner, who reviews per-client contributing factors — missed reports, open incidents, SLA warnings — and schedules internal check-ins before any client-facing outreach.
SLA forecast shows elevated breach probability for a client with four open incidents. The delivery lead reassigns capacity, resolves two items same-day, and creates a risk record documenting root cause. After Refresh on Dashboard → Predictive, breach probability drops and the intervention is auditable in the operational modules.
An agency managing eighty clients cannot review every account weekly. The director uses portfolio signals on Dashboard → Predictive to identify twelve accounts in declining or critical trend bands. Account owners receive a prioritized list with health scores, contributing factors, and recommended actions — stable and growing segments receive lighter scheduled touchpoints.
Common predictive intelligence issues
| Problem | Cause | Solution |
|---|---|---|
| Module locked or missing from navigation | Plan lacks ai_predictive_intelligence feature on current subscription | Upgrade to Professional via Settings → Billing or contact sales for Enterprise |
| All clients show unknown or low confidence | Insufficient historical activity across the portfolio | Ensure reports, incidents, monitoring data, and SLA assignments exist for active clients |
| Stale scores after major data changes | Forecasts refresh on nightly schedule by default | Click Refresh on Dashboard → Predictive after bulk imports or incident closures |
| Client missing from forecasts | New client without activity history or archived status | Confirm client status is active; allow time for data across at least one historical window |
| Revenue forecast empty | Profitability data unavailable on plan or not configured | Configure client revenue in profitability module; verify plan includes profitability feature |
| Health score differs between dashboard and client page | Scores recalculate on different refresh schedules | Use Refresh on both views for consistency after significant changes |
| API returns 403 on predictive endpoints | Plan missing ai_predictive_intelligence or API key lacks predictive.read scope | Verify plan tier and create a new API key with predictive.read scope in Settings → API |
