North American asset manager uses Duco to significantly reduce reconciliation rule building time
ASSET MANAGER
THE CHALLENGE
The client, a large North American asset manager with nearly $1t AUM, needed to perform reconciliations on complex and frequently messy data from various counterparties in structured, unstructured and semi-structured formats as part of its day-to-day business and regulatory requirements.
As this data is vital for operations, hundreds of people in middle- and back-office operations were focused on extracting, normalising, enriching and reconciling it, as well as exporting it for reporting and archiving purposes. These operations take hundreds of hours, and due to their complexity frequently end up in manual processes that add lag, opacity and ultimately risk to every day management.
If these processes fail, bad data proliferating across the business creates chaos and has the potential for significant financial, regulatory, and reputational damage for the firm.
THE SOLUTION
Using Duco’s agentic rule builder, which leverages Amazon Bedrock and LangGraph, the client has seen significant reductions on time spent building rules in their automation and reconciliation processes. This is significant when there can be as many as 92 rules for their most complex data transformation and matching processes.
The tool enables any user to build rules in Duco’s no-code, Natural Rule Language from plain-English prompts.
The client has been pleased with this reduction in build time, as it speeds overall time to market for new reconciliations.