cdp-edge 1.13.0 → 1.14.0

package/server-edge-tracker/schema-fraud.sql

@@ -0,0 +1,90 @@
+-- Schema Fraud Detection — CDP Edge Quantum Tier
+-- Versão: 1.0
+-- Data: 9 de Abril de 2026
+-- Fase 4 Enterprise-Level
+
+-- TABELA: Sinais de Fraude por Evento
+CREATE TABLE IF NOT EXISTS fraud_signals (
+  id              INTEGER PRIMARY KEY AUTOINCREMENT,
+  detected_at     TEXT NOT NULL DEFAULT (datetime('now')),
+
+  -- Identificadores do evento
+  ip_address      TEXT,
+  fingerprint     TEXT,
+  user_id         TEXT,
+  email_hash      TEXT,             -- SHA256, sem PII
+  event_name      TEXT,
+  event_id        TEXT,
+
+  -- Score e decisão
+  fraud_score     INTEGER NOT NULL, -- 0-100
+  action_taken    TEXT NOT NULL,    -- 'allowed', 'flagged', 'dropped'
+  reasons         TEXT NOT NULL,    -- JSON array: ["ip_velocity_high", "datacenter_ip", ...]
+
+  -- Contexto
+  ip_country      TEXT,
+  ip_asn          TEXT,
+  user_agent      TEXT,
+  bot_score       INTEGER,
+  velocity_1h     INTEGER,          -- Eventos deste IP na última 1h
+  velocity_1m     INTEGER,          -- Eventos deste IP no último 1min
+
+  -- Resultado (preenchido depois se lead converter)
+  was_real_user   INTEGER           -- 1 = comprou depois (falso positivo), 0 = fraude confirmada
+);
+
+-- TABELA: Alertas Agregados (quando IP/fingerprint atinge threshold repetidamente)
+CREATE TABLE IF NOT EXISTS fraud_alerts (
+  id              INTEGER PRIMARY KEY AUTOINCREMENT,
+  created_at      TEXT NOT NULL DEFAULT (datetime('now')),
+  updated_at      TEXT NOT NULL DEFAULT (datetime('now')),
+
+  alert_type      TEXT NOT NULL,    -- 'ip_attack', 'fingerprint_abuse', 'bot_network'
+  entity_type     TEXT NOT NULL,    -- 'ip', 'fingerprint', 'asn'
+  entity_value    TEXT NOT NULL,    -- O IP, fingerprint ou ASN em questão
+
+  -- Métricas do ataque
+  events_total    INTEGER NOT NULL DEFAULT 0,
+  events_dropped  INTEGER NOT NULL DEFAULT 0,
+  peak_score      INTEGER NOT NULL DEFAULT 0,
+  first_seen      TEXT NOT NULL DEFAULT (datetime('now')),
+  last_seen       TEXT NOT NULL DEFAULT (datetime('now')),
+
+  -- Status
+  is_blocked      INTEGER NOT NULL DEFAULT 0,  -- 1 = está no KV blocklist
+  blocked_at      TEXT,
+  block_expires   TEXT,                          -- NULL = permanente
+  resolved_at     TEXT,
+  resolved_by     TEXT,                          -- 'auto', 'manual'
+
+  -- Contexto
+  top_reasons     TEXT,             -- JSON: razões mais comuns
+  sample_ips      TEXT              -- JSON: amostra de IPs relacionados (para redes de bots)
+);
+
+-- Índices
+CREATE INDEX IF NOT EXISTS idx_fraud_signals_ip      ON fraud_signals(ip_address);
+CREATE INDEX IF NOT EXISTS idx_fraud_signals_fp      ON fraud_signals(fingerprint);
+CREATE INDEX IF NOT EXISTS idx_fraud_signals_score   ON fraud_signals(fraud_score);
+CREATE INDEX IF NOT EXISTS idx_fraud_signals_action  ON fraud_signals(action_taken);
+CREATE INDEX IF NOT EXISTS idx_fraud_signals_date    ON fraud_signals(detected_at);
+CREATE INDEX IF NOT EXISTS idx_fraud_alerts_entity   ON fraud_alerts(entity_type, entity_value);
+CREATE INDEX IF NOT EXISTS idx_fraud_alerts_blocked  ON fraud_alerts(is_blocked);
+CREATE INDEX IF NOT EXISTS idx_fraud_alerts_date     ON fraud_alerts(created_at);
+
+-- VIEW: Dashboard de fraude (últimas 24h)
+CREATE VIEW IF NOT EXISTS v_fraud_dashboard AS
+SELECT
+  COUNT(*) AS total_events_checked,
+  SUM(CASE WHEN action_taken = 'dropped' THEN 1 ELSE 0 END) AS events_dropped,
+  SUM(CASE WHEN action_taken = 'flagged' THEN 1 ELSE 0 END) AS events_flagged,
+  SUM(CASE WHEN action_taken = 'allowed' THEN 1 ELSE 0 END) AS events_allowed,
+  ROUND(
+    CAST(SUM(CASE WHEN action_taken = 'dropped' THEN 1 ELSE 0 END) AS REAL) /
+    NULLIF(COUNT(*), 0) * 100
+  , 2) AS fraud_rate_pct,
+  AVG(fraud_score) AS avg_fraud_score,
+  MAX(fraud_score) AS peak_fraud_score,
+  COUNT(DISTINCT ip_address) AS unique_ips_flagged
+FROM fraud_signals
+WHERE detected_at >= datetime('now', '-24 hours');

package/server-edge-tracker/schema-segmentation.sql

@@ -0,0 +1,219 @@
+-- Schema de Segmentação Dinâmica ML — CDP Edge Quantum Tier
+-- Versão: 1.0
+-- Data: 9 de Abril de 2026
+
+-- TABELA PRINCIPAL: Segmentos ML
+CREATE TABLE IF NOT EXISTS ml_segments (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  cluster_id INTEGER NOT NULL,
+  cluster_name TEXT NOT NULL,                      -- Nome descritivo gerado por ML
+  clustering_algorithm TEXT NOT NULL,                 -- 'kmeans', 'dbscan', 'hierarchical'
+  client_vertical TEXT,                                -- 'curso-online', 'ecommerce', 'saaS'
+  created_at TEXT DEFAULT (datetime('now')),
+  updated_at TEXT DEFAULT (datetime('now')),
+
+  -- Estatísticas do cluster
+  size INTEGER NOT NULL,                                 -- Número de leads no cluster
+  percentage REAL NOT NULL,                              -- % do total (ex: 0.25 = 25%)
+
+  -- Características médias (centróides)
+  avg_ltv REAL,                                        -- LTV médio do cluster
+  avg_ltv_class REAL,                                   -- 0=Low, 1=Medium, 2=High (média)
+  avg_behavior_score REAL,                               -- Engajamento médio (0-100)
+  avg_engagement_score REAL,                             -- Interações médias (0-100)
+  avg_intention_level REAL,                               -- Intenção de compra média (0-100)
+  avg_days_since_lead REAL,                              -- Recência média em dias
+  avg_hour_of_day REAL,                                  -- Hora média do dia (0-23)
+  avg_is_weekend REAL,                                   -- % que acessa no fim de semana (0-1)
+  avg_is_business_hours REAL,                              -- % em horário comercial (0-1)
+  avg_bot_score REAL,                                    -- % humano médio (0-100, 100=100% humano)
+
+  -- Características dominantes (top features por cluster)
+  dominant_countries TEXT,                                -- JSON array: ["BR", "US", "AR"]
+  dominant_states TEXT,                                   -- JSON array: ["SP", "RJ", "MG"]
+  dominant_cities TEXT,                                   -- JSON array: ["São Paulo", "Rio de Janeiro"]
+  dominant_timezones TEXT,                                 -- JSON array: ["America/Sao_Paulo", "America/New_York"]
+  dominant_utm_sources TEXT,                              -- JSON array: ["facebook", "google", "tiktok"]
+  dominant_utm_mediums TEXT,                             -- JSON array: ["cpc", "organic", "social"]
+  dominant_features TEXT,                                   -- JSON array: ["ltv", "behavior_score", "engagement_score"]
+
+  -- Métricas de qualidade do clustering
+  silhouette_score REAL,                                   -- Coesão vs separação (-1 a 1, > 0.5 = bom)
+  cohesion REAL,                                        -- Similaridade média intra-cluster
+  separation REAL,                                       -- Distância média inter-cluster
+  inertia REAL,                                          -- Soma dos quadrados das distâncias (menor = melhor)
+
+  -- Recomendações automáticas geradas por ML
+  action_recommendations TEXT,                          -- JSON array: ["Priorizar remarketing", "Aumentar bid", ...]
+  bid_recommendations TEXT,                              -- JSON array: [{"adset_id": "123", "recommended_bid": "R$ 18.50"}, ...]
+  campaign_recommendations TEXT,                           -- JSON array: [{"creative": "VSL A", "audience": "Segment 0"}, ...]
+
+  -- Metadados de controle
+  is_active INTEGER DEFAULT 1,                            -- 0 = arquivado, 1 = ativo
+  min_data_points INTEGER,                                -- Mínimo de leads para formar cluster
+  epsilon REAL,                                         -- Para DBSCAN (distância máxima)
+  min_samples INTEGER,                                     -- Para DBSCAN (pontos mínimos para cluster)
+  max_depth INTEGER                                       -- Para Hierarchical (profundidade máxima)
+);
+
+-- Índices para performance
+CREATE INDEX IF NOT EXISTS idx_ml_segments_id ON ml_segments(id);
+CREATE INDEX IF NOT EXISTS idx_ml_segments_cluster ON ml_segments(cluster_id);
+CREATE INDEX IF NOT EXISTS idx_ml_segments_algorithm ON ml_segments(clustering_algorithm);
+CREATE INDEX IF NOT EXISTS idx_ml_segments_created ON ml_segments(created_at);
+CREATE INDEX IF NOT EXISTS idx_ml_segments_active ON ml_segments(is_active);
+CREATE INDEX IF NOT EXISTS idx_ml_segments_vertical ON ml_segments(client_vertical);
+
+-- TABELA DE ASSOCIAÇÃO: Segment Members (Quem está em cada segmento)
+CREATE TABLE IF NOT EXISTS ml_segment_members (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  lead_id TEXT NOT NULL,                                  -- ID da tabela leads
+  cluster_id INTEGER NOT NULL,                              -- ID do cluster (ml_segments.id)
+  clustering_algorithm TEXT NOT NULL,                 -- Mesmo algoritmo usado para criar o cluster
+  confidence REAL NOT NULL,                               -- 0-1 (quão perto do centroide)
+  distance_to_centroid REAL,                              -- Distância Euclidiana ao centroide
+  updated_at TEXT DEFAULT (datetime('now')),
+
+  -- Metadados para rastreamento
+  assigned_at TEXT DEFAULT (datetime('now')),              -- Quando foi atribuído ao segmento
+  is_outlier INTEGER DEFAULT 0,                              -- 0 = normal, 1 = outlier (DBSCAN)
+  outlier_reason TEXT,                                     -- "behavior_score too high", "unusual geo", etc.
+
+  -- Características do lead no momento da atribuição
+  lead_ltv REAL,                                         -- LTV do lead no momento
+  lead_ltv_class REAL,                                   -- Classe de LTV do lead
+  lead_behavior_score REAL,                               -- Behavior score no momento
+  lead_engagement_score REAL,                             -- Engagement score no momento
+  lead_intention_level REAL,                               -- Intention level no momento
+  lead_days_since_lead REAL,                              -- Dias desde lead no momento
+  lead_hour_of_day REAL,                                  -- Hora do dia no momento
+  lead_is_weekend INTEGER,                                -- Se era fim de semana (0/1)
+  lead_is_business_hours INTEGER,                          -- Se era horário comercial (0/1)
+  lead_country TEXT,                                      -- País do lead
+  lead_state TEXT,                                       -- Estado do lead
+  lead_city TEXT,                                        -- Cidade do lead
+  lead_utm_source TEXT,                                  -- UTM source do lead
+  lead_utm_medium TEXT,                                   -- UTM medium do lead
+
+  -- Chave composta para evitar duplicatas
+  UNIQUE(lead_id, cluster_id, clustering_algorithm)
+);
+
+-- Índices para performance
+CREATE INDEX IF NOT EXISTS idx_ml_segment_members_lead ON ml_segment_members(lead_id);
+CREATE INDEX IF NOT EXISTS idx_ml_segment_members_cluster ON ml_segment_members(cluster_id);
+CREATE INDEX IF NOT EXISTS idx_ml_segment_members_algorithm ON ml_segment_members(clustering_algorithm);
+CREATE INDEX IF NOT EXISTS idx_ml_segment_members_confidence ON ml_segment_members(confidence);
+CREATE INDEX IF NOT EXISTS idx_ml_segment_members_outlier ON ml_segment_members(is_outlier);
+CREATE INDEX IF NOT EXISTS idx_ml_segment_members_assigned ON ml_segment_members(assigned_at);
+
+-- TABELA DE HISTÓRICO DE CLUSTERING (Auditoria)
+CREATE TABLE IF NOT EXISTS ml_clustering_history (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  clustering_id INTEGER NOT NULL,                          -- ID em ml_segments
+  started_at TEXT NOT NULL,                               -- Quando iniciou o clustering
+  completed_at TEXT,                                     -- Quando terminou
+  algorithm TEXT NOT NULL,                                 -- 'kmeans', 'dbscan', 'hierarchical'
+  n_leads_processed INTEGER NOT NULL,                        -- Quantos leads foram processados
+  n_clusters_created INTEGER NOT NULL,                        -- Quantos clusters foram criados
+  total_duration_ms INTEGER NOT NULL,                      -- Duração total em milissegundos
+  workers_ai_neurons_used INTEGER,                         -- Neurônios usados (para billing)
+  status TEXT NOT NULL,                                    -- 'completed', 'failed', 'timeout'
+  error_message TEXT,                                       -- Se falhou, motivo do erro
+  parameters TEXT NOT NULL,                                 -- Parâmetros usados (JSON)
+  results_summary TEXT,                                    -- Resumo dos resultados (JSON)
+  created_at TEXT DEFAULT (datetime('now'))
+);
+
+-- Índices para auditoria
+CREATE INDEX IF NOT EXISTS idx_ml_clustering_history_clustering ON ml_clustering_history(clustering_id);
+CREATE INDEX IF NOT EXISTS idx_ml_clustering_history_started ON ml_clustering_history(started_at);
+CREATE INDEX IF NOT EXISTS idx_ml_clustering_history_algorithm ON ml_clustering_history(algorithm);
+CREATE INDEX IF NOT EXISTS idx_ml_clustering_history_status ON ml_clustering_history(status);
+
+-- TABELA DE FEATURE IMPORTANCE (Otimização Futura)
+CREATE TABLE IF NOT EXISTS ml_feature_importance (
+  id INTEGER PRIMARY KEY AUTOINCREMENT,
+  feature_name TEXT NOT NULL,                             -- 'ltv', 'behavior_score', etc.
+  importance_score REAL NOT NULL,                           -- 0-1 (quão importante para clustering)
+  clustering_algorithm TEXT NOT NULL,                 -- Algoritmo usado para calcular
+  variance_explained REAL,                                 -- % da variância explicada
+  created_at TEXT DEFAULT (datetime('now')),
+  updated_at TEXT DEFAULT (datetime('now'))
+);
+
+-- Índices
+CREATE INDEX IF NOT EXISTS idx_ml_feature_importance_feature ON ml_feature_importance(feature_name);
+CREATE INDEX IF NOT EXISTS idx_ml_feature_importance_algorithm ON ml_feature_importance(clustering_algorithm);
+
+-- VIEW: Segmentos Ativos com Estatísticas
+CREATE VIEW IF NOT EXISTS v_active_segments AS
+SELECT
+  ms.id,
+  ms.cluster_id,
+  ms.cluster_name,
+  ms.clustering_algorithm,
+  ms.client_vertical,
+  ms.size,
+  ms.percentage,
+  ms.avg_ltv,
+  ms.avg_ltv_class,
+  ms.avg_behavior_score,
+  ms.avg_engagement_score,
+  ms.avg_intention_level,
+  ms.avg_days_since_lead,
+  ms.dominant_countries,
+  ms.dominant_states,
+  ms.dominant_cities,
+  ms.dominant_utm_sources,
+  ms.dominant_features,
+  ms.silhouette_score,
+  ms.cohesion,
+  ms.separation,
+  ms.inertia,
+  ms.action_recommendations,
+  ms.bid_recommendations,
+  ms.campaign_recommendations,
+  COUNT(msm.id) as member_count
+FROM ml_segments ms
+LEFT JOIN ml_segment_members msm ON msm.cluster_id = ms.id
+WHERE ms.is_active = 1
+GROUP BY ms.id
+ORDER BY ms.created_at DESC;
+
+
+-- VIEW: Membros de Segmentos (enriquecidos com dados de leads)
+CREATE VIEW IF NOT EXISTS v_segment_members_enriched AS
+SELECT
+  msm.id,
+  msm.lead_id,
+  msm.cluster_id,
+  msm.clustering_algorithm,
+  msm.confidence,
+  msm.distance_to_centroid,
+  msm.is_outlier,
+  msm.outlier_reason,
+  msm.assigned_at,
+  msm.lead_ltv,
+  msm.lead_ltv_class,
+  msm.lead_behavior_score,
+  msm.lead_engagement_score,
+  msm.lead_intention_level,
+  msm.lead_days_since_lead,
+  msm.lead_hour_of_day,
+  msm.lead_is_weekend,
+  msm.lead_is_business_hours,
+  l.email,
+  l.first_name,
+  l.last_name,
+  l.city,
+  l.state,
+  l.country,
+  l.utm_source,
+  l.utm_medium,
+  l.utm_campaign,
+  l.created_at as lead_created_at,
+  ms.cluster_name
+FROM ml_segment_members msm
+INNER JOIN leads l ON msm.lead_id = l.id
+ORDER BY msm.confidence DESC, msm.assigned_at DESC;