What changed
Duco, an agentic Operations platform for financial services, has rebuilt its AI stack using Amazon Bedrock and Amazon Bedrock AgentCore. This shift allows the platform to handle financial data processing more efficiently while ensuring data isolation for each financial institution.
What this means in practice
Duco processes 20 billion transactions monthly for over 200 clients, including 7 of the top 20 banks and 10 of the top 20 asset managers. By adopting Amazon Bedrock, Duco gains access to a range of foundation models from leading AI companies through a single API, enabling it to select the most suitable model for each task. The use of Amazon Bedrock AgentCore allows Duco to run agents at scale, reducing deployment time and onboarding costs. The platform operates as a multi-tenant SaaS offering with strict tenant isolation, ensuring that each financial institution's data remains completely separate from others.
Limitations and availability
The transition to Amazon Bedrock and AgentCore was driven by the need to address the limitations of traditional machine learning models, which required extensive historical data for training and struggled with novel scenarios. Duco's existing agent infrastructure was built on LangChain and LangGraph, and AgentCore is fully compatible with these frameworks, allowing most existing code to continue working without modification. The platform's multi-tenant architecture ensures that each tenant has dedicated AWS services, including databases, storage, and encryption keys, with no shared data infrastructure across clients. The supplied announcement does not specify the exact release date or availability of the updated platform.
