Overview
Key Capabilities: • Multi-agent orchestration: Supervisor, Profile, Transaction Intelligence, Document, Screening/OSINT, Knowledge and Communications agents working together for end-to-end SOW investigations. • Bedrock AgentCore: Secure identity-aware agent-to-agent/service communication. • Semantic knowledge: Bedrock Knowledge Base and S3 vectors • Enterprise security: Cognito, KMS, IAM least-privilege, Guardrails. • High availability: Multi-AZ deployment with automatic failover. • Real-time processing: SSE streaming and WebSockets. • MCP integration: AgentCore Gateway for KYC, core banking, document stores, World-Check, Google Search, case management (e.g. PEGA), and Teams/Email.
Included AWS Services:
- Amazon Bedrock (Models, Guardrails, Knowledge Base, Evals)
- Amazon AgentCore (Runtime, Memory, Observability, Identity)
- LangGraph
- Strands SDK, MCP Protocols
- S3 Vector, Cognito, Aurora PostgreSQL Serverless, Lambda, ECR
Highlights
- 65% Reduction in Manual Triage Time: Automated data extraction across required enterprise systems and structured wealth analysis significantly accelerates case handling and investigation turnaround. 75% Automation Coverage: AI-driven agents manage the majority of SOW investigation steps—including data gathering, plausibility assessment, report creation and preliminary recommendations—while human reviewers focus on final decisions and complex or high-risk scenarios.
- Enhanced Compliance: Automated audit trails record all actions, data consulted, and rationale applied, supporting AML/KYC obligations and providing a complete, reviewable investigation history. Risk Mitigation: Pattern-based wealth analysis, alignment of customer documents with observed financial behaviour, and comprehensive screening improve verification accuracy and reduce missed risk indicators.
- Improved Analyst Productivity: Analysts spend less time on repetitive data collection and document handling, enabling greater focus on judgement-based assessments and quality oversight. Consistent Quality: Standardised assessment criteria ensure uniform SOW evaluation across cases, reducing human variability and strengthening regulatory defensibility.
Details
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Pricing
Custom pricing options
How can we make this page better?
Legal
Content disclaimer
Resources
Vendor resources
Support
Vendor support
Get in Touch:
Yun Zhi Lin Partner, Deloitte Cloud & Engineering yzlin@deloitte.com.au