Using artificial intelligence in operational order management

Operational order management forms the most critical interface between a company and its customers. Its core lies in the execution of customer orders. At this interface, customer expectations, an overwhelming flow of information, and supply chain orchestration converge.

The requirements on both people and systems are high: decisions must be made under extreme time pressure, impacting the entire supply chain. Based on demand from customers’ supply chains, actions are triggered throughout distribution, inventory management, planning, production, procurement and beyond.

The foundation for this is the end-to-end digitization of all processes, along with the automation of standardized workflows and decisions. Here, Artificial Intelligence (AI) contributes to making complex decisions in a fraction of the time, leveraging vast amounts of information where human decision-making reaches its limits and is prone to errors and delays.

That’s why we see the use of AI agents in this area as one of the most important application cases in supply chain management, which we would like to present below.

Developments in recent years have shown that manufacturing companies are facing rising costs for raw materials, energy, and wages.[1] At the same time, the current economic situation brings additional uncertainty in market demand. Many industries are also dealing with overcapacity,[2] leading to a strong buyer’s market.

As a result, high service levels are essential to retain customers: short delivery times, reliable deliveries, and flexibility in order changes are critical.

Communication should be fast and frictionless. While there are different speeds in terms of information exchange, the goal is the full automation of interfaces between systems on the customer and supplier sides.

Given rising customer expectations, uncertain demand, and increasing costs, companies face several challenges:

• High working capital: Inventory levels must be optimized to reduce costs without compromising material availability and losing customers.
• Rising production costs: Increased input costs and smaller lot sizes require better machine utilization to keep costs under control.
• Ensuring service levels: Diverse service level agreements must be met to prove reliability in a buyer’s market. Transparency and proactive planning are crucial.
• Fast, reliable order processing: Established habits around ordering times, order changes, and goodwill arrangements between customers and suppliers often result in individualized, manual processes, increased effort, and delays.
• Decision-making transparency: Isolated IT systems and rudimentary digital workflows require manual, time-consuming data preparation and hinder analysis.

We see the goal of AI agents as supporting order management in efficiently, accurately, and quickly fulfilling customer demands. Smart and innovative solutions are used to support three key functions (see Figure 1):

Create real-time transparency
AI agents analyze complex data sets and generate forecasts to detect critical bottlenecks early. Well-known examples include tracking material flows and aligning demand forecasts. Based on real-time data, AI agents predict deviations in plans (e.g., delayed deliveries, increased sales volumes, reduced production output) and assess the severity of supply chain disruptions.

Make effective delivery decisions
In bottleneck situations, trade-off decisions must be made because not all customer agreements can be met.
AI agents consider various information sources, such as service level agreements, current delivery performance levels, potential contractual penalties, extra costs, and expected incoming orders. Using these models, customer orders are objectively prioritized and fulfilled.

Orchestrate the supply chain
The decisions made to fulfill customer orders require actions throughout the supply chain. AI agents support both internal and external workflow automation, e.g., triggering orders, adjusting production plans, and optimizing transport routes. A major application area is also automating customer interaction using Natural Language Processing (NLP).

By using AI agents, companies can respond more agilely to market changes and, for example, capitalize on revenue opportunities. They enable data-driven, transparent, and objective decisions that effectively meet customer requirements and minimize supply chain costs. Automating customer processes speeds up execution and reduces the error rate.

A successful example of AI at the customer-supplier interface is the collaboration between Unilever and Walmart Mexico — Unilever’s largest customer outside the U.S. As early as 2009, a joint program for collaborative planning, forecasting, and replenishment was launched. In 2022, “Sky” was introduced — an AI-powered system for analysis, planning, and orchestration of a cross-company supply chain.

Walmart provides the system with daily sales data by SKU and store. Unilever supplies promotional plans for the next three to four months. An algorithm considers historical sales data, price elasticity, current trends, events, and other parameters. This results in 3,1 million forecast combinations and 12,5 billion mathematical calculations per day, ultimately generating up to 20 million Unilever product orders for all Walmart locations and synchronizing Unilever’s supply chain down to material purchasing.

In the pilot (covering ~30% of Unilever’s business in Mexico), a 98% fill rate and 98% on-shelf availability were achieved. Forecast accuracy improved, increasing supply chain agility and resilience. Unilever estimates manual efforts could be reduced by 30% through automation. The model is now being rolled out to its 30 key customers worldwide, starting in the UK and USA.[3][4]

The Unilever-Walmart example shows a very advanced use of AI and is the result of years of process, data, and capability development. Companies should develop their own vision for AI-supported order management based on their current challenges and strategic direction.

We recommend an iterative approach to building operational order management capabilities. Depending on a company’s maturity, the initial challenges often involve structuring transaction data (e.g., historical orders, inventory, demand plans) and measuring key order processing KPIs (e.g., delivery reliability, order lead times, forecast accuracy). This provides an initial assessment of fulfillment capability based on digitally available data. These foundations then enable the digitalization and automation of simple workflows and customer interactions.

The next steps depend on the defined target vision. In our view, it’s important to build capabilities for transparency, decision-making, and orchestration in parallel (see Figure 2).

Alongside developing order management through these maturity levels, we believe that understanding the role of order management within the supply chain organization is a key success factor. That’s why we emphasize involving all stakeholders across the supply chain — e.g., sales, planning, production, procurement, suppliers — to jointly initiate a change process, build trust, and realign processes and organization.

• The role of customer service is evolving into a holistic, operational supply chain management function with greater decision-making authority and organizational significance.
• Building AI support in order management means shifting decision-making from subjective stakeholder influence toward objective criteria, clear goals, and defined service level agreements.
• As processes become automated, manual tasks shift from standardized processing toward handling exceptions and managing yet-to-be-modeled business cases.
• Building AI-powered decision models in order management is a peak innovation achievement that requires continuous learning and development. The available technical solutions are rapidly evolving. We therefore recommend setting iterative development goals — small successes help build trust.

At TenglerConsulting, we place customer requirements at the center of our efforts to optimize supply chains and processes in supply chain management. We see operational order management as the control center of a holistic supply chain. With our experience from numerous transformation projects, we understand the interests of diverse stakeholders and support companies through processual and organizational changes in order management.

Sources:
[1] Bundesministerium für Wirtschaft und Klimaschutz (DE), 2025: https://www.bmwk.de/Redaktion/DE/Schlaglichter-der-Wirtschaftspolitik/2025/02/04-jahreswirtschaftsbericht.html [accessed: 15 April 2025].
[2] Europäische Kommission, 2025: https://ec.europa.eu/commission/presscorner/detail/de/ip_25_335 [accessed: 15 April 2025].
[3] Unilever, 2024: https://www.unilever.com/news/news-search/2024/utilising-ai-to-redefine-the-future-of-customer-connectivity/ [accessed: 15 April 2025]
[4] Zero100, 2024: https://zero100.com/unilever-global-supply-chain-reboot/ [accessed: 15 April 2025]

In our series “AI in Supply Chain Planning”, we describe our top use cases of Artificial Intelligence in the supply chain. Previous use cases and other supply chain trends can be found in our insights.