SYSPRO launched its Torque artificial intelligence platform on September 2, a system designed to deploy AI agents that can spot operational problems, recommend fixes, and execute approved actions directly inside ERP, inventory, and warehouse systems. The move targets a shift in enterprise software: from AI that answers questions to AI that performs work, with distributors-including companies generating more than $2 billion in revenue-already testing the platform in a controlled rollout that began in August.
The platform entered controlled availability with manufacturers and distributors in food and beverage, industrial equipment, and fabricated metals. SYSPRO did not name the participating distributors. The company said Torque provides deeper integration with its own ERP software but is built to work across different ERP systems, warehouse technology, and cloud, on-premises, or hybrid environments.
How the AI agents operate
Torque pulls data from orders, inventory levels, supplier records, and operating schedules to identify issues and determine possible responses. Companies set the guardrails. They can require human approval before an agent takes any action or authorize it to run repeatable workflows automatically. Every decision gets logged along with the business rules and data the system used, creating an audit trail that shows how the agent reached and executed its conclusions.
Employees describe the work they want an agent to perform using plain-language instructions. Torque then builds the agent without requiring the employee to write code. That design could let operations teams apply AI to specific processes-inventory management, procurement, customer service, planning, finance-without waiting on software development resources.
Knowledge graph and system connections
SYSPRO said Torque incorporates a manufacturing knowledge graph built on nearly five decades of industry practices. The platform can connect with manufacturing execution systems, warehouse hardware, and other operational data sources. Companies in the controlled rollout are working with SYSPRO to define which decisions Torque should recommend, which actions agents can execute, and how results should be measured.
Pricing follows a usage-based model. Customers can start with a single workflow and add agents as their needs expand. The company described the approach as a way to begin small and scale based on what proves effective inside a specific operation.
Why this matters for operations professionals
The test for operations teams isn't whether AI agents can work in a limited pilot. It's whether they can move into routine operating processes while maintaining controls over the decisions they make and the actions they take. The participation of distributors with more than $2 billion in revenue gives SYSPRO a proving ground in larger, more complex distribution operations. For operations managers building the business case, the platform's logging and approval architecture addresses a central tension: giving agents enough authority to reduce manual work without surrendering oversight of decisions that affect inventory, supplier commitments, and customer fulfillment. Professionals evaluating where AI fits into their AI for Operations strategy can use the Torque model as a reference point for how agentic systems handle the handoff between recommendation and execution.
Your membership also unlocks:
