Duco: IDP Software Vendor
On This Page
Enterprise data automation platform for capital markets firms, with agentic AI workspace and AI-native document preparation for post-trade operations.
Overview
Duco provides an enterprise data automation platform for capital markets and financial institutions, built around a core argument: legacy OCR and IDP tools are engineered for static formats and break the moment data shifts. Its answer is an AI-native stack that adapts in real time, targeting the format variability and reconciliation costs that define post-trade operations.
In November 2025, Duco announced Agentic Workspace, scheduled for general availability in Q1 2026, positioning the platform as "the trusted control layer in an agentic world." The capability combines data automation, exception management, and agentic collaboration inside a single auditable environment - operations teams monitor, adjust, and work alongside AI agents in real time rather than handing off to black-box automation.
At Sibos, Duco extended that strategy into document ingestion with two additions to its Adaptive Intelligent Document Processing (AIDP) layer: the Agentic Rule Builder and expanded AI-native data preparation. Availability status for the Sibos announcements - whether generally available or in preview - was not disclosed in source material.
Duco also owns Metamaze, a Belgium-based IDP vendor acquired in February 2024, known for proprietary deep learning that requires 50% fewer training documents than comparable approaches.
How Duco Processes Documents
Duco's document processing is built on its AIDP layer, which the company positions against traditional OCR on the grounds that rule-based extraction degrades when input formats change. The Sibos announcements added two capabilities to this foundation.
Flexible ingestion extends the range of input formats AIDP can handle without reconfiguration. Low-training extraction is designed to reduce setup time when onboarding new document types - a recurring friction point in post-trade environments where counterparty formats vary widely. No benchmark figures comparing setup time or extraction accuracy against prior workflows or named competitors were disclosed.
The Agentic Rule Builder sits on top of the data preparation layer. Users write plain-English prompts; the system converts them into Natural Rule Language (NRL) rules that operations staff - not developers - can validate, edit, and apply. Consistent changes can be propagated across multiple rules simultaneously, targeting the handoff bottleneck between business logic owners and technical implementers rather than competing on raw extraction accuracy.
As Justin Hingorani, Managing Director of AI and Data Prep at Duco, framed the competitive position: "OCR and IDP were built for static formats. They break the moment data shifts. Our AI-native approach adapts in real time, making it the only sustainable path forward."
No pricing, customer adoption figures, or third-party benchmark data were available in source material. The competitive framing is directional pending independent validation.
Use Cases
Capital Markets and Post-Trade Operations
Post-trade reconciliation is the primary pressure point Duco addresses. Format variability across counterparties is high, reconciliation errors are costly, and operations teams - not developers - typically own the business logic. The Agentic Rule Builder targets this directly: plain-English rule authoring removes the developer dependency for rule creation and propagation, while the auditable Agentic Workspace gives operations staff real-time visibility into what automated agents are doing and why.
Exception management and workflow optimization sit within the same environment, so escalations from automated processing to human review happen inside a single interface rather than across disconnected systems.
Financial Institution Automation
For broader financial institution deployments, Duco's governance architecture is the differentiating claim: human oversight, explainability, and the ability to adjust automated behavior in real time. This positions the platform for regulated environments where straight-through processing must coexist with audit requirements and the ability to demonstrate control over automated decisions.
Technical Specifications
| Feature | Details |
|---|---|
| Platform Type | Enterprise data automation with agentic AI capabilities |
| Target Market | Capital markets firms and financial institutions |
| Deployment | Cloud-based; Agentic Workspace GA targeted Q1 2026 |
| Document Processing | Adaptive Intelligent Document Processing (AIDP) layer |
| Rule Authoring | Agentic Rule Builder - plain-English to Natural Rule Language (NRL), no-code |
| Data Preparation | Flexible ingestion (broad format support), low-training extraction for new document types |
| Governance | Full oversight, explainability, human control, auditable environment |
| Integration | Interoperability with existing systems |
| Availability (Sibos features) | Not disclosed - GA vs. preview status unconfirmed |
Resources
- Website: https://du.co
- Agentic Workspace announcement: Finextra, November 2025
- Sibos product launch: Duco press release, du.co
- Third-party coverage: Intelligent Document Processing Weekly Recap #173
Sources
2025-11 [press release: Agentic Workspace | finextra.com] Duco announces Agentic Workspace for Q1 2026 GA, combining data automation, exception management, and agentic collaboration (https://www.finextra.com/pressarticle/108005/duco-introduces-agentic-workspace)2025-07 [press release: Sibos launch | du.co] Duco launches Agentic Rule Builder and expanded AIDP capabilities at Sibos; availability status not disclosed (https://du.co/duco-launches-next-gen-ai-automation-at-sibos/)2025-07 [third_party: IDP Weekly Recap #173 | intelligentdocumentprocessing.com] Third-party coverage of Duco's Sibos AI-native data prep and Agentic Rule Builder announcements (https://www.intelligentdocumentprocessing.com/idp-news-weekly-recap-173/)
Company Information
London, United Kingdom
Web: https://du.co
Duco acquired Metamaze in February 2024, adding a Belgium-based IDP specialist with proprietary deep learning technology to its post-trade automation stack.
For context on the agentic AI direction Duco is pursuing, see the Agentic Document Processing capability page and the Agentic Document Processing guide.