Manhattan Associates updated its supply chain software this year exclusively with AI enhancements. In doing so, the company is making it clear that the supply chain will become increasingly dependent on AI for speed and precision. These AI innovations are inseparable from the cloud. For that reason, Manhattan is steadily nudging its on-premises customers toward cloud adoption, although still cautiously for now.
For background information on Manhattan Associates and its offerings, see: Manhattan Associates provides supply chain software, is it more than a fancy name?
Manhattan Associates is consistently guiding its customers towards the Manhattan Active Platform, the cloud-native environment on which it runs its supply chain software. CEO Eric Clark, who leads the company from February 2025 on, emphasizes that Manhattan wants customers to make the transition at their own pace. Still, the message is unmistakable: the next major upgrade for on-premises deployments will most likely be a migration to the cloud. Once there, the old cycle of major updates every five to ten years disappears, because in the cloud, all solutions are continuously updated.
This marks a notable shift. When Manhattan Associates was founded, warehouses and supply chains were entirely on-premises, and long upgrade cycles were normal. That time is definitely behind us, according to the company.
The power of the unified platform
Manhattan encourages customers to adopt the Manhattan Active Platform. Three key features distinguish this platform: it is API-first, unified, and cloud-native. Thanks to this unification, companies can see in real time which products are entering and leaving the warehouse. This is information that competitors often do not have access to. That visibility translates directly into more accurate forecasting and better inventory decisions, powered by actual data rather than assumptions. According to Manhattan, this tight integration between planning and execution is a major differentiator.
Product Leader Brian Kinsella explains how unification creates new opportunities. When a trailer arrives for receiving, for example, the system pre-calculates the optimal storage locations for all items and automatically selects the receiving door that minimizes put-away travel distance. This works because the yard management system can call the put-away API within the warehouse management system.
Manhattan’s software runs on Google Cloud. Both parties learn from each other in this collaboration. Manhattan can count on Google’s expertise in developing software and AI solutions, while Google, in turn, collects practical examples. This is actually an important source of inspiration for future developments.
Optimization through AI
The company’s biggest announcement is the introduction of agentic AI. CTO Sanjeev Siotia presented a vision of a future in which hundreds of autonomous AI agents help teams gain insights faster, coordinate complex internal operations, and execute tasks independently under human supervision.
Manhattan will provide roughly 20 pre-built agents, but expects customers to build many of their own. The prevailing opinion is that companies must get to work themselves, because every company has its own specific use cases where a general agent will fall short. Manhattan does provide support for this, for example, through the development of an agent-building AI agent in the Manhattan Agent Foundry. This is again based on Google technology: Google Agentspace. All Manhattan agents will be available on the AI Agent Marketplace.
Specific applications that are already available:
Knowledge Assist Agent: A conversational AI assistant that lets users ask questions in natural language rather than navigating complex menus. Instead of navigating complex menus, employees can simply ask questions such as, “Is there excess labor capacity that I can use to get the picking done on time?” The system analyzes the situation and provides concrete answers, including the names of available employees. It has already handled hundreds of thousands of conversations and is reshaping how people interact with supply chain software.
Fulfillment Spotlight: A feature that pinpoints where bottlenecks originate in supply chain planning. Companies now get concrete insights into what they can improve, where they previously had to rely on guesswork.
Service Insight: This tool analyzes the customer experience in after-sales care. This part of the purchase deserves attention. Companies can measure customer satisfaction using the NPS (Net Promoter Score), where customers can use a ten-point scale to indicate how likely they are to recommend the company to friends and family. Manhattan notes that this score is assessed equally by the purchase and the after-sales experience. Service Insight provides metrics and suggestions for improvement to optimize that experience.
Dynamic by Assignment: A practical example of AI optimization: upon trailer arrival, the system determines optimal storage locations in advance and assigns the receiving door that yields the shortest put-away distance.
Unified Business Plan: This new functionality in Manhattan Active Supply Chain Planning goes beyond traditional forecasting. The system not only predicts what inventory will move through the network, but also how many employees are needed per function in the distribution center. Forecasts can be made for 6 weeks ahead, up to 52 weeks. These projections are based on concrete data such as promotions, seasonal influences, and causality, rather than simplistic sales growth percentages.
The challenge of in-store connectivity
A persistent challenge discussed during the event relates to cloud connectivity in warehouses and stores. Many companies have migrated enterprise applications to the cloud and rolled them out at headquarters, but extending that rollout to distributed stores and warehouses is proving much slower.
As a result, the connectivity footprint of stores today is still far from what it should be for running cloud apps and AI components live. This issue persists even after the rollout. On the one hand, it is influenced by a store’s location, which, for example, depends on the infrastructure available within a shopping center. On the other hand, it is influenced by network failures, as Manhattan notes that retailers with hundreds of stores will notice every day that several locations experience failures.
This reality complicates Manhattan’s cloud-first vision. However, the company is addressing the issue by investing heavily in redundancy measures that keep cloud applications running despite connectivity disruptions. The goal is to ensure that the customer’s shopping experience remains unaffected.
Conclusion
Manhattan Associates is resolutely steering towards a cloud-only future. For on-premises customers, the message is unmistakable: the next upgrade means a migration to the cloud. Those who don’t will miss out on the AI-driven future of supply chain operations. Something that the orchestration of all incoming data should benefit from.
For this article, we spoke with members of Manhattan Associates’ leadership team. From left to right: CEO Eric Clark, SVP Brian Kinsella, CTO Sanjeev Siotia, and Executive Vice President Bob Howell.