The first available capability is a natural language assistant that allows users to query Nasdaq Calypso data, documentation and other information. It operates within the platform’s agentic AI environment, which provides a contained and governed space for deploying AI capabilities across capital markets and treasury operations.
The Night Shift You Never See: How AI Agents Are Changing Capital Markets Operations
And how Nasdaq Calypso plans to bring Agentic AI to users
Key Takeaways
- AI agents can perform defined operational work while keeping people accountable for outcomes.
- Structured, auditable workflows make capital markets a strong proving ground for agentic AI.
- Governance, authoritative data and human oversight are essential to trusted adoption.
- Nasdaq Calypso is building toward governed AI capabilities within established workflows and controls.
Picture the start of your day: You pour a coffee, sit down and open your screen. Before you begin, AI agents have already completed defined overnight work within established workflows. Reconciliation breaks have been analyzed, settlements at risk of failing have been flagged and exceptions have been prioritized for review.
The work has not been taken out of your hands. Instead, the analysis and preparation are ready, allowing you to begin with the issues that need your judgment rather than the steps required to find them. Your priorities are clear.
This is becoming possible for capital markets operations through advances in AI and agentic workflows.
AI agents represent a new source of intelligent capacity. More than another tool or dashboard, they can carry out defined work when called by a person, workflow, schedule or event, helping teams across front, middle and back offices focus their time on judgment, priorities and strategy.
How close are we to this new reality? We announced our first capability, available today, which helps users investigate and interpret information within established workflows. From that starting point, Nasdaq Calypso plans to support the future development of governed AI agents that could help users work more efficiently and effectively over time, within established workflows and controls.
Myth, Reality or Just a Better Chatbot?
Image
A chatbot is conversational and human-led: it answers, explains and guides when a person asks. An AI agent goes further by carrying out defined work when called by a person, workflow, schedule or event (or pre-programmed to execute tasks on a daily basis). An agentic worker represents a more advanced operating model in which AI can assess context, determine that action may be needed and initiate defined work within established permissions, controls and governance boundaries. The distinction is not how many AI capabilities can be orchestrated, but what starts the work.
For capital markets firms, the relevant near-term opportunity is the AI agent. Agents can analyze information, coordinate defined steps and produce recommendations within clear boundaries while people direct or oversee the work and remain involved at defined points.
The supporting technologies, governance models and operating patterns are beginning to take shape. Agent orchestration, workflow automation and governance layers can support bounded use cases, but deployment depends on how institutions redesign work, establish controls and determine where human review or approval is required.
The challenge for many firms is trust, governance and scalability. Nasdaq Calypso is seeking to address these priorities by preserving its deterministic, mission-critical transaction processing foundation while augmenting user productivity and operational efficiency through AI-driven assistance and automation. AI can inform and accelerate decisions, but the platform’s core trading, risk, collateral and post-trade processes are designed to remain controlled, auditable and deterministic.
Why Capital Markets Technology Is the Proving Ground
Capital markets operations is a strong proving ground for AI because the work is already structured around workflows, policies, approvals and data.
Deterministic processes are designed to be ring-fenced in the application, and user and agent steps are designed to be systematically audited. Reconciliation, settlement and payment exceptions, margin and collateral, risk investigation and regulatory processing are governed by rules, leave an audit trail and have a clear notion of a correct outcome. Those conditions create a foundation for agents to support defined work without weakening control.
This is where the opportunity and the challenge meet. Capital markets operations teams have every reason to show progress on AI, move faster and reduce friction, yet the same workflows that stand to gain the most are also mission critical and heavily regulated. The book of record has to stay dependable and the calculations auditable. The firms that advance are the ones that find a way to capture the efficiency without loosening control, and that is the problem the right platform should help solve.
Nasdaq Calypso: From a System of Record to a System of Action
This is the agentic shift Nasdaq Calypso is seeking to accelerate. The platform that institutions already trust as their system of record is extending toward an intelligent platform of action, helping users investigate issues, interpret context, prioritize work and support decisions within existing workflows and controls.
The principles behind this are deliberate and intentional. AI here is additive. The core stays as clients rely on it: deterministic, auditable and predictable, now with a governed foundation for introducing AI capabilities across the workflows clients already rely upon.
In practice, that means future potential for agents working inside the workflows people already run.
The following examples illustrate the types of operational workflows where governed AI agents could create value over time. They represent future potential, illustrating what AI agents could do, rather than a definitive product roadmap.
- Valuation, risk and P&L: A move is analyzed down to its drivers, with a drafted explanation ready for review.
- Reconciliation: A break is detected, the pattern behind it is read against history and a remediation is proposed.
- Settlement and payments: Trades and payments most likely to fail are identified early and surfaced to support timely intervention.
- Exception handling: The day's exceptions arrive already ranked and reasoned, so work begins with the items that matter most.
- Workflow management: A start-of-day action plan provides a personalized operational briefing that highlights critical exceptions, risks, priorities and recommended actions, which could help enable faster issue resolution and improved operational efficiency.
- Limits and compliance: Exposures could be analyzed on an ongoing basis to help identify potential risks earlier and provide recommendations to traders, risk managers and operations teams.
Image
Trust Is the Anchor
What helps make agentic AI viable in a regulated environment is governance, auditability, thoughtful deployment and robust data protection. Authoritative data is fundamental to that, emphasizing the importance of agents operating inside a system itself, rather than bolted on as a separate AI layer. This is critical to controls and data lineage.
Human oversight is another important dimension of trusted AI design. Agents can help answer and explain, progress to recommending actions and act only at defined approval points. Efficient and effective tools codify the role of human judgment and decision-making. With calibrated guardrails, adoption can move at the pace each institution’s governance permits.
Over time, trusted AI strategies will also depend on openness and control. Firms will likely need to retain flexibility across internal, vendor-provided and third-party agents as evolving industry standards mature. The goal is to extend intelligent workflows through shared understanding while preserving governance and control.
Helping Accelerate Agentic AI
AI agents and advances in agentic technology can create new capacity across capital markets operations while preserving human accountability.
A reconciliation lead can work alongside AI agents that analyze breaks and prepare recommended actions. A collateral manager may draw on both human and agent capacity to manage the same book of work. A junior risk analyst can partner with an agent to proactively analyze risk exposures and time series, then equip a human manager with clearer business priorities. People remain accountable for the outcome, while the division of work is calibrated to each workflow and the controls.
Which brings us back to the start of the day. The work is not simply completed outside human view. Defined analysis and preparation are ready for review, so people can begin with the issues that require judgment. The capabilities are taking shape now, and capital markets operations provide a practical proving ground. With end-to-end workflows and controls used by financial institutions worldwide, Nasdaq Calypso is positioned to help firms deploy governed AI within the processes they already rely on.
Information set forth in this post contains forward-looking statements, including statements regarding the planned development and future capabilities of AI agents for Nasdaq Calypso. Nasdaq cautions readers that any forward-looking information is not a guarantee of future performance and that actual results could differ from those contained in the forward-looking information. Forward-looking statements can be identified by words such as “will,” “plans,” “could,” “may,” “seeking,” and “believe,” and other words and terms of similar meaning. Forward-looking statements involve a number of risks, uncertainties or other factors beyond Nasdaq’s control. Nasdaq Calypso is a product of Nasdaq’s Financial Technology business and is operationally independent and distinct from The Nasdaq Stock Market, LLC.
© 2026 Nasdaq, Inc. The Nasdaq logo and the Nasdaq ‘ribbon’ logo are the registered and unregistered trademarks, or service marks, of Nasdaq, Inc. in the U.S. and other countries. All other trademarks, including Sibos, are the property of their respective owners. All rights reserved. This communication and the content found by following any link herein are being provided to you by Nasdaq Financial Technology, a business of Nasdaq, Inc. and certain of its subsidiaries (collectively, “Nasdaq”), for informational purposes only. Nothing herein shall constitute a recommendation, solicitation, invitation, inducement, promotion, or offer for the purchase or sale of any investment product, nor shall this material be construed in any way as investment, legal, or tax advice, or as a recommendation, reference, or endorsement by Nasdaq. Nasdaq makes no representation or warranty with respect to this communication or such content and expressly disclaims any implied warranty under law. At the time of publication, the information herein was believed to be accurate; however, such information is subject to change without notice. This information is not directed or intended for distribution to, or use by, any citizen or resident of, or otherwise located in, any jurisdiction where such distribution or use would be contrary to any law or regulation or which would subject Nasdaq to any registration or licensing requirements or any other liability within such jurisdiction. By reviewing this material, you acknowledge that neither Nasdaq nor any of its third-party providers shall under any circumstance be liable for any lost profits or lost opportunity, direct, indirect, special, consequential, incidental, or punitive damages whatsoever, even if Nasdaq or its third-party providers have been advised of the possibility of such damages.
© 2026 Nasdaq, Inc. All rights reserved.
FAQs
The environment applies operational boundaries, live oversight and strict sandboxing with no external data retention. It is designed to keep AI agents operating within an institution’s perimeter and policies while allowing clients to retain control of their data.
The operating environment will allow clients to access Nasdaq Calypso agents and connect their own proprietary AI infrastructure through an integrated layer built on the Model Context Protocol, an emerging industry standard for linking AI agents to software.