Duco Rolls Out Agentic Operations Platform for Post-Trade Finance

By Milan Rojan

Duco has entered the agentic AI space with a purpose-built post-trade operations platform, as financial institutions faced mounting pressure from compressing settlement windows, rising transaction volumes and a generational shift in operational workforces.

The London-based FinTech has built the offering on its existing infrastructure, which processed 20 billion transactions monthly for more than 200 clients, including seven of the top 20 global banks and ten of the top 20 asset managers. Clients included CIBC Mellon, ING and Man Group.

Rather than layering artificial intelligence onto legacy systems, the firm has unbundled its infrastructure into hundreds of discrete agent-accessible capabilities covering reconciliation, data preparation, exception management, audit trails and document creation, accessible via Model Context Protocol (MCP). The move has reflected wider efforts across financial services to embed automation directly into post-trade workflows.

Early deployments have pointed to tangible efficiency gains. A reconciliation process that previously required two days was completed in approximately four hours, with agents accounting for roughly 20 minutes and the remainder allocated to human review.

“For more than a decade, our clients have trusted Duco to reconcile the most complex data in capital markets. They are now telling us that agents will run a meaningful share of post-trade Operations within three years.  Duco is assembling the most powerful set of tools and context ever built for Operations to enable this,” said Christian Nentwich, CEO and co-founder of Duco.

Alongside the launch, Duco has opened a second wave of its Pacesetters programme, a cohort of financial institutions given early access to new capabilities and direct input into product development. Ten firms had already deployed Duco agents in production environments at the time.

The system incorporated human-in-the-loop validation for consequential actions, role-based permissions consistent with existing user controls and comprehensive audit trails recording every agent action, provisions designed to support regulatory requirements.