How abo7adeed/omerta.ai works
Omerta.ai is a financial-crime investigation and banking platform with customer and compliance web experiences. Customers use the portal to manage accounts and transfers; compliance staff monitor risk, investigate cases, and review evidence and reports. The README documents double-entry banking, risk controls, network analysis, support, and multi-agent forensics. The graph preserves those product branches and the sampled Agentic RAG internals; unsampled routes and services are represented only at supported directory-level granularity.
Main entry points: main.py FastAPI Gateway · nodes.py RAG Orchestration · models.py PostgreSQL · transfer_service.py Transfers and Ledger
Components 24
Customer and Staff Apps
- Customer Portal
frontend/src/pages/customer/SendMoneyPage.tsx - Compliance Console
frontend/src/App.tsx
API and Identity
- FastAPI Gateway
apps/api/main.py - JWT Identity
infrastructure/security/jwt_auth.py
Banking and Risk
- Transfers and Ledger
domain/services/transfer_service.py - Risk Assessment
infrastructure/risk/ml_provider.py - Banking Services
domain/services
Investigation and RAG
- Investigation Workflow
apps/investigator/account_workflow.py - Specialist Agents
domain/investigation/specialist_agents.py - Investigation Scoring
domain/investigation/scoring.py - RAG API
apps/api/v1/agentic_rag.py - RAG Graph
apps/agentic_rag/graph.py - RAG Orchestration
apps/agentic_rag/nodes.py - Conversation Memory
domain/agentic_rag/memory.py - RAG Data Tools
domain/agentic_rag/tools/database_tools.py - Graph Tools
domain/agentic_rag/tools/graph_tools.py - Evidence Visualization
domain/agentic_rag/visualization.py - Document Ingestion
infrastructure/knowledge/ingest.py
Persistence and Integrations
- PostgreSQL
infrastructure/database/models.py - Neo4j Graph
infrastructure/neo4j/client.py - MCP Tool Servers
mcp_servers/graph_server/server.py
Other components
- Customer
- Compliance Staff
- LLM Provider
Connections 28
- Customer → Customer Portal: uses
- Compliance Staff → Compliance Console: uses
- Customer Portal → FastAPI Gateway: requests
- Compliance Console → FastAPI Gateway: requests
- FastAPI Gateway → JWT Identity: enforces identity
apps/api/main.py - FastAPI Gateway → Transfers and Ledger: handles transfers
- FastAPI Gateway → Banking Services: routes banking operations
apps/api/main.py - Transfers and Ledger → Risk Assessment: evaluates risk
domain/services/transfer_service.py - Transfers and Ledger → PostgreSQL: records ledger
domain/services/transfer_service.py - FastAPI Gateway → Investigation Workflow: runs investigations
apps/api/main.py - Investigation Workflow → Specialist Agents: uses specialists
apps/investigator/account_workflow.py - Specialist Agents → Investigation Scoring: applies scoring
domain/investigation/specialist_agents.py - FastAPI Gateway → RAG API: serves RAG
apps/api/main.py - RAG API → RAG Graph: invokes graph
apps/api/v1/agentic_rag.py - RAG Graph → RAG Orchestration: dispatches nodes
apps/agentic_rag/graph.py - RAG Orchestration → RAG Data Tools: queries records
apps/agentic_rag/nodes.py - RAG Orchestration → Conversation Memory: uses memory
apps/agentic_rag/nodes.py - RAG Orchestration → Graph Tools: queries graph
apps/agentic_rag/nodes.py - RAG Orchestration → Evidence Visualization: renders evidence
apps/agentic_rag/nodes.py - RAG Orchestration → Document Ingestion: uses knowledge corpus
apps/agentic_rag/nodes.py - RAG Data Tools → PostgreSQL: reads records
domain/agentic_rag/tools/database_tools.py - Graph Tools → Neo4j Graph: queries graph
domain/agentic_rag/tools/graph_tools.py - Document Ingestion → PostgreSQL: stores indexed chunks
infrastructure/knowledge/ingest.py - MCP Tool Servers → PostgreSQL: serves banking tools
- MCP Tool Servers → Neo4j Graph: serves graph tools
- Investigation Workflow → LLM Provider: uses model
apps/investigator/account_workflow.py - RAG Orchestration → LLM Provider: uses model
apps/agentic_rag/nodes.py - FastAPI Gateway → PostgreSQL: persists results
apps/api/main.py