Authors: Brett Vogelsang and Kedar Prabhudesai
We are excited to introduce an AI Copilot for the SAS Strategic Supply Chain Optimization Model. We designed it to democratize access to advanced modeling capabilities. With intuitive, conversational interfaces, team members at all levels of technical expertise can now easily run complex optimizations, interpret results, and perform scenario analyses.
This innovation comes at a critical time. According to Gartner, leading supply chain organizations are leveraging AI to optimize their processes at more than twice the rate of their lower-performing counterparts. This widening “AI advantage gap” is reshaping the competitive landscape. It distinguishes future-ready organizations from those at risk of falling behind. As supply chains grow increasingly complex and volatile, AI is no longer just a tool for operational efficiency. It's a fundamental business differentiator.
At SAS, we recognized this shift early. Our approach to embedding Generative AI (GenAI) with the SAS Strategic Supply Chain Optimization Model is built on a simple yet powerful premise. Even the most sophisticated supply chain optimization tools provide value only when they are accessible to those making critical operational decisions.
Traditional supply chain functions involve diverse teams with specialized expertise. It runs the gamut from Sales teams focused on customer demand and Logistics personnel managing transportation networks to Warehousing staff optimizing inventory placement and Production teams balancing manufacturing capacity. These teams have deep domain expertise but often limited technical capability to harness the full power of SAS Analytics platforms independently. The complexity of statistical programming, model configuration, and data interpretation creates an accessibility barrier that forces these professionals to either depend on centralized analytics resources or make decisions without optimization insights. They must do all this while managing their demanding day-to-day operational responsibilities. This persistent gap between supply chain expertise and analytics proficiency ultimately leads to underutilized technology investments. This also leads to missed opportunities for operational excellence.
Historical background
This work builds on previous endeavors from the SAS Applied AI and Modeling Division. They developed a reusable LLM-SAS integration framework through the creation of GenAI Assistants for warehouse optimization and recipe optimization. These projects provided key insights into mitigating mathematical hallucinations and integrating SAS with GenAI to deliver meaningful value, empowering users who may not have deep technical expertise in data analysis but possess the business acumen to interpret results and drive decision-making.
Key capabilities unlocked with AI Copilot
By integrating GenAI with the SAS Strategic Supply Chain Optimization Model, we empower non-technical users to run optimizations, review resulting data changes, and generate “what-if” scenarios based on tacit business knowledge using a natural language interface. Here are some key capabilities of the AI Copilot.
Model execution
Customers can execute the optimization model using natural language. This means that non-technical users, whether in sales, logistics, warehousing, or production, can tap into advanced AI-driven insights simply by asking questions in plain language.
Data summarization
Overall Network: In Figure 1, customers can view a high-level summary of the entire network.
Customers: Customers can review a customer summary to understand how their customers were impacted after running the optimization model. Figure 2 demonstrates this.
Distribution Centers: Customers can review a distribution center summary to understand how their distribution centers were impacted after running the optimization model. See Figure 3.
Suppliers: If customers want to know how their suppliers were impacted after running the optimization model, they can review a supplier summary. Figure 4 displays this.
Plants: If customers want to know how their manufacturing plants were impacted after running the optimization model, they can review a plant summary as shown in Figure 5.
Scenario analysis
The real power of the AI Copilot for supply chain optimization lies in its ability to make it easy to execute scenarios modelling the impact of changes in the supply chain network. Four possible scenarios are available in the first release of the AI Copilot.
1. Location Shutdown (Plant, Supplier, Distribution Center): If a customer plans to shut down a plant for scheduled maintenance, they can ask the AI Copilot how that will impact the overall network, as seen in Figure 6.
2. Demand Changes: If a customer is planning to run a promotion for a product and expects demand to increase for a product by a certain multiple, they can understand how that will impact the overall network by asking the AI Copilot, as seen in Figure 7.
3. Manufacturing plant line time availability changes: A customer may be interested in increasing capacity for a specific set of lines within a manufacturing plant. By using the AI Copilot, they can understand how that will impact the overall network.
4. Updating the Global Variables for the Optimization Model: Instead of manipulating the data themselves, customers could ask the AI Copilot to modify either the objective function or the penalty values in the global variables table.
Once customers have reviewed all of their scenarios of interest, they can compare the high-level metrics from each scenario to decide which one to promote. See Figure 8.
Value add with AI Copilot
Integrating our AI Copilot with Supply Chain Optimization creates transformative value across three critical dimensions. These reshape how organizations leverage optimization insights.
- Accelerating Time-to-Insight: The AI Copilot reduces time-to-insight from days to minutes. This is done by automating data preparation, streamlining analysis workflows, and translating complex outputs into clear, actionable narratives. This enables teams to focus on strategic decision-making rather than data wrangling.
- Democratizing Optimization Capabilities: The conversational interface lowers the technical barrier to entry. This allows business analysts, operations managers, and other non-technical stakeholders to independently run complex scenarios without requiring specialized data science expertise. By using Natural Language, the reliable results from SAS Analytics become accessible and useful to technical and non-technical teams across the enterprise.
- Enhancing Decision-Making at Every Level: AI Copilot supports executive-level decision-making in S&OP meetings by delivering concise summaries and scenario comparisons. At the same time, it provides tailored insights for functional teams across sales, logistics, warehousing, and production. This multi-level support ensures that optimization insights flow seamlessly across the organization, fostering a unified approach to supply chain excellence. One that extends beyond isolated pockets of analytical expertise.
AI Copilot summary
Gartner's research highlights a widening AI advantage gap in supply chain management. Leading organizations are adopting AI at more than twice the rate of their peers. This competitive divide has direct implications for operational resilience and market responsiveness.
Our AI Copilot for Supply Chain Optimization bridges this gap by transforming how teams interact with complex optimization tools. Moreover, with the natural language interface, anyone, regardless of technical expertise, can run sophisticated models. They can translate complex outputs into actionable insights and explore scenarios in a conversational manner. This democratization of advanced analytics goes beyond operational efficiency. It creates a strategic advantage by distributing decision-making capabilities across the organization and accelerating responses to market shifts.
The question isn't whether AI will transform supply chain optimization—it already has. The real question is: How quickly will your organization adapt to this new paradigm?
If you're interested in learning more about how we can integrate the AI Copilot for Supply Chain Optimization into your workflow, please reach out to request a demo.
Learn More | SAS Models Request a Demo | SAS Models Read More | Warehouse Optimization Use Case
Kedar Prabhudesai
Kedar Prabhudesai is a Senior Data Scientist in the SAS Applied AI and Modeling Division. He has over a decade of experience in machine learning and data science. His work spans diverse areas including time-series forecasting, computer vision, and, more recently, generative AI. Kedar holds a Ph.D. and an M.S. in Electrical and Computer Engineering from Duke University.
