Duco, the AI-powered data automation company, has announced the launch of “Agentic Workspace,” a new collaboration layer within the Duco Platform designed to integrate human expertise with autonomous AI agents. Planned for general availability in Q1 2026, the workspace aims to provide financial institutions with a transparent, auditable environment where operations teams can build, maintain, and monitor intelligent agents without losing oversight.
As firms increasingly adopt agentic AI, the challenge of maintaining control over automated systems has become a critical operational hurdle. The Agentic Workspace is designed to shift the user experience from manual process construction to intent-based interaction, a fundamental evolution in how users engage with data automation technology, according to the company.
From No-Code to Natural Language
“While Duco has always been an advanced no-code platform where users build and manage their own processes, we’re now moving that concept forward,” Duco’s Chief Product Officer, James Maxfield, tells TradingTech Insight. “This shift completely changes how people interact with the platform. Duco was already great at narrowing down thousands of exceptions to the handful that matter, but our new agentic layer supercharges that ability.
Instead of manually running processes, users can just query a chat interface. It does the heavy lifting – ingesting information, building and running processes and surfacing exceptions – presenting everything in a way that is far more intuitive.”
This capability addresses a persistent pain point in capital markets operations: the volume of data noise. “Too many teams are still stuck manually hunting for needles in haystacks to make sense of their data,” notes Maxfield. “Our agentic framework is the solution to that operational bottleneck.”
Deconstructing the Legacy “Jenga” Stack
One of the primary use cases for the new workspace is tackling technical debt, specifically the “black box” legacy files and End User Computing (EUC) tools that plague financial institutions.
“Consider the common issue of technical debt: obscure XML files or hard-coded legacy logic left behind by employees who retired years ago,” explains Maxfield. “Now, we can take that ‘black box’ – whether it’s a PDF, a macro, or an old file – and have our agent reconstruct its function directly inside Duco.”
Maxfield highlights the compliance benefits of this approach, noting that it allows firms to retire fragile, unmonitored tools. “This is hugely impactful. It allows organizations to finally move away from the thousands of risky macros and EUCs that regulators hate, migrating them into a safe, controlled environment.”
However, the transition away from legacy systems often stalls due to fear of disruption, observes Maxfield. “Many firms are terrified to touch their legacy tech. It’s like Jenga – move one block and you don’t know what will collapse. That fear is one of the biggest blockers to transformation, but our agentic capability removes that risk. The agent decomposes, understands, and rebuilds processes quickly and safely. It acts as the technical expert, proposing the build, while the ‘human in the loop’ remains essential to approve each step.”
Governance and Velocity
To ensure this level of automation does not compromise security, Duco has designed the agents to follow the existing permission structures of the firm.
“Crucially, the agent operates within the same guardrails as any employee,” Maxfield asserts. “It interacts with the platform exactly like a human user, subject to the same permissions and governance controls – simply an intelligence performing the actions rather than a person.”
Underpinning this innovation is Duco’s cloud-native architecture, which enables the platform to leverage broader advancements in generative AI to accelerate its own development cycle.
“Duco’s cloud-native heritage is a massive advantage. We run on AWS, meaning we directly benefit from their massive investment in Bedrock,” says Maxfield. “This allows our engineering teams to leverage the latest technology and build sophisticated agents in two-week cycles.”
Looking ahead to the 2026 launch and beyond, Maxfield has set an aggressive pace for updates. “My goal is to make this our cadence for next year: delivering fresh updates to the agentic workspace every two weeks.”
Subscribe to our newsletter
