How Ninth Wave built AI-powered open finance onboarding on Amazon Bedrock
Learn how Ninth Wave built Compass, a multi-agent AI onboarding assistant on Amazon Bedrock AgentCore that validates bank APIs against Financial Data Exchange (FDX) standards, scores compliance, and compresses open finance onboarding from weeks to minutes while meeting SOC 2 and PCI DSS requirements.
Ninth Wave, a company specializing in secure data connectivity between financial institutions and third-party applications, has developed Compass, an AI-powered onboarding assistant built on Amazon Bedrock AgentCore. Compass aims to enhance the onboarding experience for banks participating in open finance by streamlining the integration process with standardized APIs.
The multi-agent architecture of Compass was designed to address the complexity of onboarding tasks while maintaining high accuracy and security standards. By dividing tasks among multiple specialized agents, Compass ensures efficient processing of API validation, field mapping, readiness scoring, and interactive Q&A. Each agent operates with its own context and instruction prompt, preventing interference between tasks and improving overall accuracy.
To provide a governed AI experience for external partners, Compass incorporates content-safety controls and audit logging while adhering to financial-services compliance standards such as SOC 2 and PCI DSS. The AI agents are grounded in the bank's own data through a tenant-scoped grounding process, ensuring that one bank's data does not compromise another bank's session.
Compass leverages Amazon Bedrock's managed multi-agent orchestration and access to multiple models, allowing the team to focus on domain logic rather than LLM infrastructure. With support for models across AWS Regions and robust security controls, including AWS Identity and Access Management (IAM), AWS Key Management Service (AWS KMS), and private networking, Compass offers a secure, efficient, and compliant solution for onboarding in the open finance ecosystem.
Written by urgent.news from AWS Machine Learning's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
