SCOUT - Statistical Client Observation & Unified Tracking

Your early warning system for GA4 anomalies across the entire STM portfolio

What is SCOUT?

SCOUT is an intelligent anomaly detection platform that monitors 50+ client GA4 properties, identifying issues before they become problems. Like a scout on reconnaissance, it's always first to spot patterns and report back with actionable intelligence.

"SCOUT found it first, so you can fix it fast."

Core Capabilities

• 🔍 Schema Discovery: Automatically maps custom events across diverse client implementations
• ⚡ Multi-Method Detection: Statistical analysis (Z-score, IQR) plus ML pattern recognition
• 📊 Cross-Client Intelligence: Learn from patterns across the entire portfolio
• 📧 Smart Alerting: Morning digests via email, critical alerts via Teams
• 🎯 Business Impact Scoring: Prioritizes anomalies by actual business impact

Project Status

Currently in implementation phase. Foundation pipeline under construction.

Tech Stack

• Data Layer: Google BigQuery (all client GA4 exports)
• Processing: Python Cloud Functions (batch daily)
• ML/Analytics: BigQuery ML, Statistical Analysis
• Alerting: SendGrid (email), Teams Webhooks
• Dashboard: Next.js + Tremor (Phase 2)

Quick Start

# Clone and setup
cd "C:\Users\Charles Blain\CascadeProjects\projects\PROJECT-058-AnomalyInsights"
python -m venv venv
venv\Scripts\activate
pip install -r requirements.txt

# Configure environment
cp .env.template .env
# Edit .env with your BigQuery project details

# Test SCOUT's schema discovery
python scripts/scout_schema_discovery.py

Documentation

• Implementation Guide: See QUICK_START.md
• Technical Architecture: See CLAUDE.local.md
• Progress Tracking: See STATE.md
• WBS & Dependencies: See .claude/wbs-planning.md

Project Metadata

• Project ID: PROJECT-058-AnomalyInsights (keeping for consistency)
• Alias: SCOUT
• Type: data-pipeline
• Created: 2025-09-25

---

SCOUT: Always watching. Always ready.