Robert Kugel's Analyst Perspectives

I recently attended Kinaxis’ annual user group meeting, Kinexions. The most important theme of the event was the trifecta of artificial intelligence (AI), generative AI (GenAI) and agents. Supply chain planning and execution software are the business software domains that are going to be major beneficiaries of AI and agentic AI automation. However, they also are going to be two of the more difficult to translate from potential to reality. Some of the basics, especially predictive analytics built on machine learning, are already available. But unlike accounting or sales which are usually built around a single or a few systems of record, supply chain planning and execution in midsize and larger organizations is built around an intricate web of systems. Since its inception, Kinaxis has focused on tackling the most complex and demanding supply chain problems, initially on sales and operations planning (S&OP), but more recently on the execution side. Its challenge now is to remain at the forefront in applying these technologies to improve the productivity, efficiency, agility and service level delivery capabilities of enterprises with the most demanding supply chain environments.

This year’s Kinexions took place at a time when supply chain disruption and the potential for AI to improve planning and execution were top of mind in the audience. The state of supply chains rarely gets attention except when shelves are bare, repair parts aren’t available, products are suddenly taking months to deliver or trade wars are erupting. The pandemic was a major source of obvious disruption and tariffs are a headline item at the moment, but the post-war ever-liberalizing trade environment began fraying in the mid ‘teens as the growth in global trade decelerated and global trade tensions began to rise (even if they were not front-page news at the time). Meanwhile, other legal or regulatory constraints such as greenhouse gas emissions have added to the challenge of finding optimal solutions in dynamic markets. This new environment has made it increasingly difficult to design products and their supply chains for achieving the lowest cost while increasing the importance of having adaptive and agile systems and processes to deal with ever-increasing disruptions. That’s why the successful deployment of AI and related technologies will have such a significant impact on supply chain operations specifically, and their broader ability to enhance enterprise performance.

With this in mind, one significant announcement at the event was the broader partnership with Databricks to enhance AI-enabled supply chain orchestration. The deeper integration of Kinaxis’ Maestro supply chain planning and execution platform, which is grounded on a data fabric element, with Databricks' Data Intelligence platform is designed to facilitate data acquisition, management and governance. (See the recent Analyst Perspective on Databricks by my colleague, Dave Menninger.) Data is the foundation for successful use of everything related to AI. However, having easily accessible, timely and reliable data is proving to be a challenge for almost all midsize and larger enterprises. ISG Research asserts that by 2027, almost all supply chain planning providers will have a dedicated data store to facilitate the integration of all data necessary for ongoing training of AI models to ensure they are performative.

The data requirements for supply chain planning and execution are especially demanding because of the cross-functional and enterprise-wide nature of the systems that support this work. Connecting data from these islands of automation at scale is a significant barrier to successful adoption of AI technology. It is a challenge that Databricks is designed to address with capabilities in data engineering, data warehousing, stream data processing, data governance, data sharing, data science and AI. The deeper collaboration aims to enable Kinaxis’ customers to unify data and accelerate AI adoption so as to improve their supply chain agility and resilience. This partnership will help organizations manage fragmented data sources more effectively, enabling faster and smarter decision-making using predictive AI, GenAI and agentic AI.

Kinaxis also recently announced a partnership with Infor, with the introduction of Kinaxis Planning One for Infor CloudSuite Industrial Enterprise, CloudSuite Automotive, and CloudSuite Aerospace and Defense. The combination is designed for discrete manufacturers that need an advanced supply chain orchestration capability. Planning One is designed expressly for midsize enterprises to address their need for a dedicated supply chain planning and execution tool that meets their cost and operational constraints. The majority of these enterprises use a mass of spreadsheets for this purpose and can benefit from using a dedicated application that allows them to do scenario planning and plan optimization while considering the financial implications of decisions. Kinaxis is pairing its software with its RapidStart implementation methodology or a similar certified partner implementation approach to achieve a low-risk, relatively shorter time to value outcome. For Kinaxis, the partnership gives it a practical approach to addressing the midsize enterprise market where Infor has a significant presence in discrete manufacturing.

AI was the central technology theme at the conference where the company announced and demonstrated numerous advances. Kinaxis had previously introduced Maestro Chat, a natural language-based generative AI tool, which is now widely in use by its customers. The new functionality performs straightforward but time-saving tasks such as the creation of customizable dashboards. Users also can ask questions in multiple natural languages and receive immediate answers to complex questions. This basic tool essentially upskills supply chain practitioners by reducing the technical skills necessary to interact with data and work with the system. The software effectively collapses decision cycles by enabling organizations to analyze scenarios, assess risks and make informed decisions without requiring advanced analytical, statistical or AI skills. In predictive AI, the company showcased easier-to-use predictive AI capabilities designed to make these capabilities more widely accessible to supply chain professionals.

Agentic AI holds significant promise because it can enable autonomous action, reducing time spent by humans on repetitive tasks and compressing task and process cycle times to enhance responsiveness and agility. It’s usually the case that a majority of decision nodes in supply chain tasks and processes follow set rules with limited or no nuance. These lend themselves to using sophisticated robotic process automation. However, the remaining are anything but cut-and-dried, requiring actors to understand multiple layers of context sensed from a set of data and actions that are sourced from scattered islands of information and execution. Those tasks, in turn, are the major source of delays in decision and execution cycles and therefore avoidable costs. Agentic AI adoption is likely to trace the familiar pattern of going from simple to sophisticated in the familiar crawl-to-run change management progression.

Kinaxis will have out-of-the-box agents for task execution available to users this quarter. These agents are designed to be context aware and able to read, modify, help visualize and analyze data using a natural language conversational interface that is familiar to anyone who has used a co-pilot or generative AI tool. In a demonstration of this capability, an individual first asks a supply-chain question (“Are there promised orders that cannot be delivered?”), and the system then asks the user to define a promised order and what “cannot be delivered” means, while providing prompts to help resolve ambiguities. The system then provides the in-context answer, and the user refines the question to zero in on the specific information they need following a set of natural language prompts. The user then asks for recommendations on how to deal with these orders and gets a detailed set of recommended options for handling the situation that users can act on. Later this year, the company will have configurable AI agents that will be able to access multiple resources, create instructions to follow, define a set of next steps, perform a sequence of tasks and work with other Kinaxis agents.

The demonstrated agentic AI capabilities are relatively simple in the context of ISG’s broader definition of this technology, but they are an important first step. The winners in the business applications AI race will be determined not just by technology but their successful adoption by customers. As has been the case with every information technology evolution, there is an often overlooked or underestimated learning curve. For that reason, the capabilities of the software provider’s system are available long before the mainstream adopters are ready to use them.

Over the past decade, the demanding environment for supply chain planning and execution has forced enterprises to take a more strategic approach to managing these processes. The need for resiliency and adaptability in a rapidly changing world economy almost always raises costs unless organizations devise ways of mitigating the impact. AI in all its forms can provide the ability to plan, react and execute with greater precision in shorter time frames to promote agility. It can also devise new ways to reduce inefficiencies in a highly complex process. I recommend that organizations with even moderately long or complex supply chains review their supply chain planning and execution processes to assess areas for improvement. As they consider how dedicated software can improve their performance, I recommend they investigate Kinaxis to determine how it might enhance resiliency and competitiveness.

Regards,

Robert Kugel