Duco has launched an agentic operations platform for post-trade financial services, built on infrastructure processing 20 billion transactions monthly.
The UK-based firm has restructured its platform into a modular agent layer built around Model Context Protocol (MCP), which exposes hundreds of discrete capabilities covering reconciliation, data preparation, data access, audit trails, exception management, and document creation. The architecture allows autonomous agents to interact with verified, deterministic toolsets rather than replacing existing matching, rules, or audit functions, a distinction the company positions as central to operational trust and accuracy.
Addressing structural pressures in post-trade
The launch targets three converging pressures in post-trade operations: shorter settlement windows, increasing transaction volumes, and changing workforce dynamics. These conditions have made manual processing harder to sustain and exposed the limitations of legacy systems in accommodating AI-driven workflows.
Early production results reported by participating firms suggest measurable efficiency gains. According to Duco, building a new reconciliation process on the platform has been reduced from two days to four hours, with approximately 20 minutes of agent runtime and the remainder allocated to human review. Additional capabilities in early deployment include auto-built workflows from raw inputs, continuous optimisation of existing processes, and accelerated exception investigation.
Alongside the platform launch, Duco has announced a 'Pacesetters' cohort, representing a group of financial services firms with early access to new capabilities as they are released, as well as direct input into product development. Ten firms are reported to be live with Duco agents in production as of the launch date. A second wave of applications has opened to qualifying financial services firms.
In addition, the company said that the platform as the operating system for post-trade in the agentic era, noting that clients have indicated agents are expected to handle a meaningful share of post-trade operations within three years. It was also added that findings from the Pacesetters cohort will be shared more broadly to accelerate industry-wide adoption.
