Industry: Service Parts

Supply chain organizations are rapidly adapting to a competitive workforce, new technologies, and a growing supply chain ecosystem. To keep pace and ensure sustainable growth, investments in technology to improve the productivity of people are increasing. Many are turning to artificial intelligence (AI), specifically generative AI, and decision intelligence tools. According to the 2023 Gartner Future of Supply Chain Survey, 40% of top performing respondents are already utilizing artificial intelligence and machine learning (AI/ML) to optimize their supply chain decisions for demand forecasting, while 31% are leveraging AI/ML for supply planning.

Not all AI/ML tools are created equal, however. Generic business-level generative AI applications are not well-suited for supply chain analysis because they lack the deep understanding of complex supply chain connections and data necessary to extract meaningful supply chain insights. However, Logility’s GenAI capability is designed specifically for supply chain use cases, providing critical answers about top products, forecast insights, demand plans, and more.

In this post, we share how the team at Logility is approaching generative AI and decision intelligence with a supply chain ontology mindset across the platform and what it means for supply chain leaders and their teams.

Generative AI Designed for Supply Chains

Generative AI is all the buzz right now with so much hype around both its clever and interesting opportunities within business, and conversely what it doesn’t do very well. Generative AI technology serves organizations well enough today within the appropriate constraints. One of the advantages of existing large language models is that they come pre-equipped with a generic “understanding” of language and the relationships between information implied by it. However, as problems move from the generic to a specific domain like supply chain, the language we use becomes both more specific and simultaneously nuanced within that domain.

Logility is investing significant energy into building an embedded supply chain ontology (a means of mapping concepts and the relations between them) into its GenAI solution to overcome this nuance challenge. Developing a key set of understandings within the Logility solution focused on how supply chain data interrelate with one another allows for more efficacy and avoids hallucinations that can happen in a more generic generative AI solution. Logility’s implementation is purpose built around the supply chain and ensures we’re giving clients and their employees the type of value they need more quickly.

From Turn Over to Tuned In

I’ve heard frequently from clients they need greater productivity out of their workforce. The challenge these organizations face is they no longer have the luxury of long training periods beside employees with thirty years of experience and expertise. This requires new hires to get up to speed instantly. We are already seeing GenAI overcome this challenge, showing productivity benefits through training, implementing forecast changes, and actionable insights.

For instance, if a new planner goes into Logility GenAI to update their forecast, they can describe the activity they want to do and our GenAI can help them navigate in real time to the appropriate page within the solution for that activity. This navigation capability effectively becomes an interactive training tool for new employees. It not only tells them where to go, but also helps them to leverage their supply chain knowledge while getting productivity gains out of software they may not have been exposed to before.

Democratize Insights

Historically, seeking answers to complex questions within a supply chain solution would require specific expertise in the tool to use it the right way to get down to the data they need. If the solution didn’t offer flexibility or openness for the particular question being asked, it might require that information to be exported from the solution into a data warehouse and analyzed with sophisticated analytics skills and tools to find the answer. Now, GenAI democratizes access to supply chain information.

With the power of natural language processing that comes with Logility GenAI you can express your inquiry in business terms without the requirement to build the working knowledge of the planning solution expected of a planner. We leverage our supply chain ontology behind the scenes to ensure the right data is efficiently retrieved, combined correctly, and displayed clearly. In addition to the initial query results, the solution transparently describes how the data was retrieved in the header of the response. Going a step further, GenAI provides additional actionable insights based on the AIs understanding of the data. This gives the client confidence in the results that are provided.

GenAI also helps those not in a supply chain role, such as finance or sales, who know how to ask for what they want, but don’t know how it structured in the platform. GenAI bridges that gap completely, democratizing insights across the entire organization and improving decision making.

See Logility GenAI Explainable Insights in Action:

Vision for the Future of Planning Productivity

In addition to the capability available in our GenAI solution today, we’re also investing in the next opportunity for improving planner productivity. We’re specifically targeting work for which it might be easy to confuse substantial activity with substantial progress.

As an example from the past, if a promotion was running, the planner might need to update the forecast across multiple products, within a given market, across the period of the promotion. Depending on the tool in use, this could require significant navigation and modifications through several areas of the forecast. Historically, a change like this might also require deep understanding of both data and the supply chain tool followed by a significant number of mouse clicks and substantial data manipulation. This sheer number of individual changes increased the likelihood of transcription errors creeping into the process.

The next opportunity within GenAI capability is to allow planners to express, in their own terms, what they need to accomplish when changing a large amount of values and to provide the description in business terms.  The interaction must include a confirmation process to validate that the request is well understood and accurate before applying the resulting changes.  If the response is not quite right, the planner can tweak the request until they can validate that the correct change would be applied. 

Not only will this functionality make changes closer to the speed of business, but it will facilitate doing so more quickly and with fewer errors originating from the sheer number of manual edits taking place. It also provides interaction opportunities for the generative AI solution to anticipate which key metrics might be impacted by the change and have those automatically incorporated into the verification step.