Player Intelligence & Data

Intelligence & RiskPhase 1

Lead: Intelligence Orchestrator Agent

player-intelligence-serviceArbitrium: In Progress

10 agents10 planned

Linear: HB-PLAY

48 permissions (8 write)Orchestrator-Workers+ Parallelization

Architecture Pattern

Intelligence Orchestrator assigns analysis tasks. Segmentation, LTV, and behavior agents run in parallel on player data. Batch operations process large datasets.

Evaluation

4 metrics defined

Error Handling

3 scenarios covered

Self-Improvement

2 human checkpoints

Phase 1: Foundation

Foundation. Every other agent depends on player segments, LTV, and churn scores. Must be built first.

Tools

Back Office APIAnalyticsLinear

Goals

Segment playersPredict LTV and churnFeed insights to all departmentsBulk user operationsMonitor agent performance

Agent Sizing Rationale

10 agents: 1 orchestrator + 3 analytics specialists (odd for segmentation voting) + 4 BI/data (including batch operations) + 2 agent governance (audit bot + performance analyst). Segmentation decisions use 3-agent consensus to avoid bias.

Player Analytics (3-agent panel)

AgentDescriptionComplexityRolesStatusActions

Segmentation AgentPerforms RFM analysis, creates behavioural cohorts, and maintains player lifecycle tags.medium

analystgenerator

LTV Prediction AgentForecasts lifetime value per player and segment using historical patterns and current behaviour.medium

analyst

Churn Prediction AgentIdentifies at-risk players with probability scores and recommends retention interventions.medium

analyst

Business Intelligence & Data (3 specialists + orchestrator)

AgentDescriptionComplexityRolesStatusActions

Reporting AgentGenerates daily, weekly, and monthly reports on all key business metrics.medium

analystgenerator

Data Pipeline AgentManages ETL processes, data quality checks, and ensures clean data flows to all agents.medium

executor

A/B Test Analysis AgentDesigns experiments, monitors statistical significance, and reports results.medium

monitoranalystgenerator

Batch Operations AgentHandles bulk user operations — batch currency adjustments, mass user tagging, batch state changes, and bulk balance corrections. Processes Finance and Operations bulk requests with audit trails.medium

executoranalyst

Intelligence Orchestrator AgentCoordinates all intelligence work, distributes insights to CRM, VIP, and Risk departments.complex

orchestratorrouter

Agent Governance (2 agents)

AgentDescriptionComplexityRolesStatusActions

Agent Audit BotTracks every AI agent action (agent ID, action type, timestamp, input/output, decision reasoning). Flags anomalies like agents approving too many high-risk operations. Generates daily/weekly agent activity reports.medium

monitoranalystgenerator

Agent Performance AnalystAnalyzes agent performance metrics: approval rate vs. human override rate, error rate (false positives/negatives), response time, throughput. Generates recommendations for agent tuning and optimization.medium

analystgenerator

Agents Used From Other Departments

These agents from other departments feed data into or are called by this department's agents.

Risk & FraudExternal

Risk Scoring Agent

Risk scores enrich player profiles and inform churn prediction.

Casino OperationsExternal

Casino Orchestrator Agent

Game activity data is a primary input for player segmentation.

SportsbookExternal

Sportsbook Orchestrator Agent

Betting activity data feeds into player analytics and LTV models.

MarketingExternal

Attribution & Analytics Agent

Attribution data from Marketing feeds into player acquisition cost analysis and channel-level LTV calculations.

RetentionExternal

Retention Orchestrator AgentChurn Intervention Agent

Churn intervention outcomes and retention campaign data feed back into predictive models, improving churn prediction accuracy.

Compliance & LegalExternal

Data Privacy Agent

Data privacy requirements constrain what player data can be collected, how long it is retained, and what models can be built from it.