OMP extends UnisonIQ with Agent Library and Agent Studio: AI agents that produce executable supply chain plans

The challenge | Bekaert | What’s new in UnisonIQ?

OMP today extended UnisonIQ, the AI framework in its Unison Planning™ platform, with the Agent Library of ready-to-deploy AI agents and Agent Studio, where companies build their own. The agents run on the optimization engines in Unison Planning™ and produce executable supply chain plans. Bekaert is among the early adopters of UnisonIQ.

Supply chain organizations face board-level pressure to show progress on AI, but many struggle to move beyond experiments. Generic AI models and decision-intelligence tools miss the industry-specific context and semantics that planning depends on, while planning tools with an AI layer on top lack reasoning depth.

UnisonIQ agents work differently: each one works from a digital twin of the customer’s own supply chain and calls on the optimization engines in Unison Planning™, so the plan changes it prepares respect real-world constraints such as shelf life, co-products, and shared capacity.

Bekaert: comparing S&OP scenarios in plain language

Bekaert, a leader in steel wire transformation and coating technologies, uses Unison Companion in the sales and operations planning (S&OP) process of its Energy Transition subdivision. Planners describe the scenario they want to test in plain language, such as adding capacity at bottlenecks while demand rises. UnisonIQ then models it, runs it, and compares it with the current plan. Bekaert is now extending Unison Companion to demand planning in its Steel Wire Solutions division, where planners use it to challenge their forecast. In parallel, Bekaert runs machine learning forecasting alongside its statistical forecasts in its Rubber Reinforcement division.

“Our planners can describe different scenarios in plain language and compare the modeled outcomes with the current plan. This helps keep our S&OP discussions focused on evaluating options and making informed decisions.”

Filip Lanckmans, Global Supply Chain Manager at Bekaert

“Most companies don’t struggle to start with AI. They struggle to scale it,” said Tom Wouters, Chief Product Officer at OMP. “An agent is only as good as the engine behind it. UnisonIQ agents run on the same optimization engines behind our customers’ plans, so planners can hand over more work as trust grows and stay in control."

What’s new in UnisonIQ

• Agent Library: Ready-to-deploy agents handle tasks across supply, demand, and data management. Rather than letting a general model estimate safety stock, an agent finds it in the digital twin and computes it with the logic in Unison Planning™. With multi-agent orchestration, several agents work through a problem together, like a team of planners.
• Agent Studio: Companies with in-house expertise can configure and build their own agents, grounded in OMP’s domain expertise rather than starting from a blank page. 
• Unison Companion orchestrates the agents: Planners ask questions in plain language. Companion decides which agents and optimization engines to run, including those from the Agent Library and Agent Studio, and returns an answer or a proposed plan change for approval. It can also draw on OMP documentation and the company’s own planning documents.
• Governance set upfront: Each organization sets its own guardrails, deciding what an agent may decide, execute, or escalate, so autonomy grows only as fast as confidence does. 
• Open architecture: UnisonIQ can connect to the broader enterprise ecosystem using AI protocols, so it extends what a company already runs.
• Built to scale: Industry context is built in, so agents work against real constraints from day one rather than stopping at a proof of concept. Each new agent