عربي
HUC logo
HUC logo
Generative AI (Large Language Models) in Supply Chain Management
Dr. Ramakrishna Yanamandra From Different Corners

Generative AI (Large Language Models) in Supply Chain Management

Activities of Supply Chain Management have changed rapidly in the recent past due to the emergence of AI and digital technologies. Many processes in supply chain became automatic, relatively more accurate and fast.

Generative AI and Large Language Models significantly contributed to this transition in supply chain from normal tools to decision-making and decision-support tools. These tools have the ability to make decisions in several areas of supply chain such as forecasting of demand, procurement and purchasing, risk identification and management, highlighting the exceptions in the system to develop a futuristic plan. Most of the manual, routine and day-to-day tasks in supply chain are automated with the application of natural-language interaction using complex models. Large language models, usually termed as LLMs have the ability to generate human-readable recommendations enabling the unfamiliar and technically weak users also to get benefitted.

 

Generative AI and LLMs can support supply chain operations broadly in decision-support, augmented forecasting and planning, procurement and supplier management and logistics planning and optimization functions. LLMs have the ability to translate plain model outcomes of supply chain operations into decision models and workflows. This provides the users across supply chain to make analytical decision for timely delivery of products.

 

Similarly, in forecasting and planning function of supply chain LLMs have the ability to integrate data from all entities and members, develop multiple-decision making models, and provide output by developing a scenario by considering the risk and uncertainty. Decisions related to inventory management are very critical in managing supply chain effectively. These decisions are linked to how well the supplier management is executed. In the manual processes, supply chain managers encounter several issues in taking timely and accurate inventory decisions by coordinating with the suppliers. However, with the application of GenAI techniques, users can execute supplier contracts, respond quickly to supplier queries, complete the proposals automatically. This will minimize the cost of supply chain operations and reduces the time taken drastically. Due to these abilities, the product availability improves across the supply chain and customer satisfaction enhances. GenAI and LLMs also have the ability to accelerate exception handling, develop optimal logistics route map, provide comparative cost statements etc. This enables supply chain users to automate some of the functions such as vendor evaluation and rating, vendor audit, maintenance of quality standards in procurement and timely acquisition of right information at any point of time.

 

GenAI and LLMs have very critical and valuable role to play in Warehouse Management System (WMS) in supply chain. There are many instances of its successful implementation such as Mecalux’s Easy WMS is integrated with a generative-AI chat interface. Through this chat, the users can send queries in simple sentences as used in an inter-personal talk. Questions such as status of orders, inventory status of a particular product, or total stock in a particular zone of a warehouse can be asked through this chat interface and obtain answers in a tabular or graph or image format. This kind of output takes couple of days to prepare manually without these technologies.

 

These technologies also enable users to modify the warehouse layouts to make them optimal and cost-effective using simple queries. For instance, a user can ask ‘arrange today’s priority orders separately’. Situations such as excess stock and stock-outs in the warehouse can be avoided through the alerts and warnings issued in the user dashboard. Reorder levels, replenishment schedules, status of safety and buffer stocks are also provided to the user well in advance to avoid uncertainty and risks.

 

Another important area of application of these technologies is in maintenance of machines and warehouses. Deviations from minor to major such as abnormal noise, sounds, unusual vibration, fluctuations in temperature and power levels, machine breakdown etc. can also be known to the user in advance through alerts. Due to these alerts, the machine downtime can be controlled resulting in reduction of cost of maintenance. Automated Guided Vehicles (AGVs) are enabled with Generative AI and LLMs to sort, place, track and retrieve the goods in warehouse. Drones are used for appropriate placement of stock in the warehouse, spoiled or damaged goods and boxes are identified by these technologies easily. Simplification of these tasks improves agility of warehouses, optimizes the space utilization, improves machine maintenance and reduces the machine downtime, and improves the overall productivity of a warehouse, which results in enhanced supply chain efficiency.

 

Industry examples of usage of Generative AI and LLMs:

 

Horizon3 system, integrated with GenAI and LLMs is used by Unilever company to get overall sales data, market and weather forecast at different global locations. This integration resulted in reduction of errors in supply chain planning by almost thirty percent for Unilever.

Retail giant, Walmart uses GenAI to analyze the trends in customer preferences, make automated decisions related to inventory improving the efficiency in its warehouses located all over the world (Menache, 2025).

 

Denmark based global logistics company Maersk, uses these technologies to make automated supplier contracts and to obtain vendor ratings and feedbacks. This automation reduced its manual contract review process time by almost fifty percent, resulting in speedy contract system and delivery of goods.

Cognizant uses these technologies to develop automated Request For Proposal (RFP), automated evaluation of tenders submitted by the vendors and selection vendors automatically.

Similarly, another global logistics company, DHL uses MySupplyChain platform integrated with Generative AI and LLMs for tracking of orders, day-to-day logistics operations, storage of documents, and to provide answers to queries related to logistics.  

 

In a similar manner, FedEx uses its own SenseAware platform to provide real-time information to the customer in tracking operations and for monitoring and identifying the deviations and disruptions in the delivery processes. 

 

Other areas of usage of these technologies is done in providing environmental data to several companies. For instance, data related to carbon levels in supply chains of various companies is provided by the LLM integrated Microsoft Cloud for Sustainability software. This data is vital for many logistics companies in view of global awareness about environment and regulatory requirements.

 

United Arab Emirates is one of the few first countries in adopting Generative AI and LLM applications in the supply chains of various sectors. It developed smart ports, smart logistics and smart road monitoring systems and smart governance through these applications. Masdar City, the sustainable city in UAE integrated AI in its supply chain operations to automate the supplier selection process.

 

Masdar City (Abu Dhabi) procurement functions used AI-based platforms to streamline supplier selection and performance tracking. Companies such as e& UAE (Formerly Etisalat), Majid-Al-Futtaim (MAF) group, Careem, Etihad Airways are some of the leading companies using these technologies.

 

References:

  • Accenture. (2024, May 6). Reinventing supply chains with generative AI. Accenture Middle East. https://www.accenture.com/ae-en/blogs/consumer-goods-services/reinventing-supply-chain-generative-ai
  • Aghaei, R., Kiaei, A. A., Boush, M., Vahidi, J., Barzegar, Z., & Rofoosheh, M. (2025). The potential of large language models in supply chain management: Advancing decision-making, efficiency, and innovation. arXiv preprint arXiv:2501.15411. https://arxiv.org/abs/2501.15411
  • Al Furqan Shipping. (2025, February 12). Generative AI in GCC freight forwarding and logistics. https://alfurqanshipping.com/generative-ai-in-gcc-freight-forwarding/
  • Complete AI Training. (2025). AI gets to work: How machine learning and GenAI are making warehouses smarter. https://completeaitraining.com/news/ai-gets-to-work-how-machine-learning-and-genai-are-making/
  • Dedicatted. (2025). AI in supply chain management and its top applications in the era of Industry 4.0. https://dedicatted.com/insights/ai-supply-chain-management-and-its-top-applications-in-the-era-of-industry-4-0
  • DHL Group. (2024). AI and automation in supply chain operations: Case study series. DHL Supply Chain Insights. https://www.dhl.com
  • DP World. (2025). Smart trade, smart ports: Digital twin and AI integration report. DP World Insights. https://www.dpworld.com
  • Galkin, A., Samchuk, G., Kopytkov, D. et al. (2025). Digital twins in logistics: a comprehensive bibliometric analysis for advancing smart cities and sustainable development. Discover Sustainability 6, 853. https://doi.org/10.1007/s43621-025-01754-0
  • McKinsey & Company. (2025). The state of AI: Global survey and generative AI adoption trends. McKinsey Insights. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
  • Mecalux. (2025). Generative AI transforms Easy WMS: New chat interface for warehouse data insights. https://www.mecalux.com/logistics-articles/generative-ai-easy-wms-mecalux
  • Menache, I. (2025). How generative AI improves supply chain management. Harvard Business Review. https://hbr.org
  • Simchi-Levi, D. (2025). Large language models for supply chain decisions. arXiv preprint arXiv:2507.21502. https://arxiv.org/abs/2507.21502
  • Wang, H., Jiang, J., Hong, L. J., & Jiang, G. (2025). LLMs for supply chain management. arXiv preprint. https://arxiv.org/abs
  • Xiros, V., Gonzalez Castro, J. M., Fernandez-Pelaez, F., Magoutas, B., & Christidis, K. (2025). Supply chain data analytics for digital twins: A comprehensive framework. Applied Sciences, 15(12), 6939. https://doi.org/10.3390/app15126939