ketteQ Launches Quintus as Agentic AI Moves Into Supply Chain Decision-Making

ketteQ has launched Quintus, an AI agent designed to operate across existing supply chain planning and ERP environments, as the industry’s AI conversation shifts from forecasting and recommendations toward systems capable of reasoning over operational problems and taking action.

Generally available as of August 12, Quintus is positioned by ketteQ as “Free-Range AI” for supply chain — the company’s term for an AI architecture that is not restricted to predetermined use cases or a single planning platform. According to ketteQ, the agent can analyze live supply chain data, generate and test scenarios, execute tasks and operate above existing systems including SAP IBP, Kinaxis, o9 and Blue Yonder without requiring customers to replace their current technology stack.

The launch arrives at an important point for supply chain technology. Generative AI has already made natural-language interaction increasingly common across enterprise software, but the more consequential question is whether AI can progress from explaining what has happened to helping organizations decide — and potentially execute — what should happen next.

For supply chain leaders, where disruptions can quickly affect inventory, production, customer commitments and revenue, that distinction matters.

From Supply Chain Copilot to Execution Layer

Much of the first generation of enterprise AI has been assistive. Users ask questions, summarize information or receive recommendations, while the underlying workflows and decisions remain largely unchanged.

Quintus is intended to push beyond that model.

A planner, for example, could ask what would happen if a supplier facility became unavailable for two weeks. Rather than retrieving an existing report, ketteQ says Quintus can analyze live information, determine which customer commitments are exposed and reason through potential responses.

The same intelligence can be surfaced outside the traditional planning organization. A CFO could examine revenue exposure from a disruption, while sales teams could check whether a customer commitment remains achievable. ketteQ says those interactions can take place through tools such as Microsoft Teams, Slack and email rather than requiring every user to work directly inside a planning application.

“Every function that has given people direct access to reasoning AI has been transformed by it. Supply chain can have that same advantage,” said Mike Landry, Founder and CEO of ketteQ. “Quintus does not just advise. It acts. And you do not have to replace a single system to get it.”

That last point may prove particularly significant. Enterprises have invested heavily in ERP and specialized planning platforms, making wholesale replacement expensive, disruptive and often strategically unnecessary. An AI layer capable of working across those investments could offer a different route to modernization.

Real-Time Scenario Analysis Becomes the Battleground

Central to Quintus is PolymatiQ, ketteQ’s patent-pending agentic solver.

The company says the technology can run thousands of constrained planning scenarios simultaneously and return an optimized response in seconds. Quintus can also generate and execute Python code dynamically as it reasons through questions rather than relying exclusively on predefined queries and workflows.

This distinction gets to one of the fundamental challenges facing AI in supply chain planning.

A conversational interface alone does not fundamentally change planning if the AI is limited to interpreting outputs that have already been calculated. The larger opportunity lies in connecting natural-language reasoning with the underlying optimization and scenario-analysis capabilities required to evaluate new conditions as they emerge.

If that model proves reliable at scale, it could change the role of planners. Instead of spending significant time assembling data, running scenarios and identifying exceptions, teams could increasingly focus on evaluating proposed responses, managing trade-offs and intervening where human judgment is required.

An AI Layer Above Existing ERP and Planning Systems

Perhaps the most strategically important aspect of the launch is ketteQ’s decision to position Quintus as technology that can operate independently of the underlying planning platform.

The company says deployments can sit above SAP IBP, Kinaxis, o9, Blue Yonder or combinations of existing enterprise systems, with implementation beginning within four to eight weeks and no rip-and-replace requirement.

That approach reflects an emerging architectural question across enterprise software: does AI ultimately become another feature embedded inside individual applications, or does it develop into an intelligence and execution layer that works across them?

For supply chain organizations, the latter model has obvious appeal. Enterprise data and workflows rarely reside in one system. ERP may hold transactions and financial information, while planning platforms, CRM applications, warehouse systems and other operational technologies each contain different pieces of the decision-making context.

An AI agent that can reason across those boundaries could potentially make the existing application landscape more useful without requiring organizations to consolidate everything onto a single platform.

But it also raises questions around integration depth, data quality and authority. The value of an AI agent depends heavily on whether it has access to accurate, timely information and whether its actions can be reconciled with the systems that remain authoritative for enterprise transactions.

Early Production Deployments Put Claims to the Test

ketteQ says Quintus is already operating in production environments at companies including Alliance Consumer Group, JCI, Mativ, NCR and Zeus, with additional deployments expected.

At Alliance Consumer Group, ketteQ reports that inventory turns increased from 2.75 to 4.0, while real-time available-to-promise information can now be delivered inside Salesforce in less than 15 seconds during customer conversations.

Stephanie Larson, Director of Supply Chain at Alliance Consumer Group, described the experience as a shift in how teams begin their working day.

“The first morning Quintus told us exactly what needed attention before we’d even logged in, it felt like the first time I used ChatGPT,” Larson said.

The production deployments are important because agentic AI in enterprise software is increasingly surrounded by ambitious claims. The meaningful test will not be whether an agent can demonstrate an impressive response in a controlled environment, but whether it can repeatedly support decisions under real operating constraints, changing data and competing business priorities.

Governance Becomes More Important as AI Starts Acting

Greater autonomy also changes the risk equation.

When AI merely summarizes information, an incorrect answer may inconvenience a user. When an AI system begins recommending or executing actions affecting inventory, production, purchasing or customer commitments, explainability and control become operational requirements.

ketteQ emphasizes that Quintus operates within a governed framework in which decisions are auditable and explainable and can be overridden by humans.

That architecture will be critical as supply chain AI moves toward execution. Organizations will need clarity over what an agent is authorized to do independently, what requires approval and how every consequential decision can be reconstructed after the fact.

The most successful agentic systems may therefore not be those offering the greatest theoretical autonomy, but those that allow businesses to determine precisely where autonomy is appropriate.

Why This Matters for Supply Chain Technology

Quintus represents a broader transition taking place across enterprise software.

ERP and supply chain systems have traditionally been built around structured transactions, predefined workflows and periodic planning cycles. AI introduces the possibility of a more dynamic operating model: continuously evaluating changing conditions, exploring alternatives and initiating responses while the underlying systems remain the enterprise systems of record.

For supply chain planning in particular, this could be significant. Supply chains operate in an environment where assumptions can become outdated within hours due to supplier disruption, demand changes, transportation constraints or production issues. The ability to reason against current conditions rather than wait for the next planning cycle could materially change how organizations respond.

There is still a substantial gap between agentic AI’s potential and proving that autonomous decision-making can be trusted across complex global supply chains. Data quality, integration, governance and human oversight remain central challenges.

But ketteQ’s launch points toward where the competition is heading. The next phase of supply chain AI is unlikely to be defined simply by which platform has the most capable chatbot. It will increasingly be defined by which systems can connect reasoning with optimization and governed execution across the enterprise technology environment already in place.

If that transition succeeds, AI may become less of a feature within supply chain software and more of a decision layer across it.