Spotfire Copilot for Supply Chain Inventory Management

Overview of Spotfire Copilot

Spotfire Copilot™ is a free, natural language extension to the Spotfire® platform. It leverages large language models (LLMs) to augment business intelligence and artificial intelligence, all in Spotfire. 

This release is custom-built to perform the following tasks in Spotfire: 

• Question-answering for Spotfire questions
• Question-answering for user-loaded documents
• Auto visualization generation and modification
• Data function generation
• Explanation of visualizations
• Data interrogation

For more information on Spotfire Copilot, check out this community article.

Application Overview

Modern global supply chains are complex ecosystems. Understanding where stock is at any given point in time is key to optimizing inventories. The Spotfire Supply Chain Inventory Management application shows an example of how stock can be analyzed over time, and provides models for ordering optimization and promotion simulation. It also provides insights into the performance of suppliers and transportation providers, and potential weather impact on deliveries. The goal is to reduce overhead costs to drive profitability.

This application is part of the Continuous Supply Chain Accelerator. For more information on the accelerator, check out this article.

Demo Walkthrough

In this demo, I’m interested in seeing if there’s a specific area where carrier performance is low - I can gauge carrier performance by seeing how many units were shipped vs. how many units were received. First, Copilot can help me figure out which particular distribution center and product we should focus on when investigating carrier issues.

Copilot has let me know that at San Jose for Intense Mocha, there’s an overall discrepancy of 3,816 units. That’s a lot! If I click on San Jose and the product Intense Mocha, the bars showing units shipped and units received in the visualization seem to show some pretty large differences across several months. 

I’m going to want to do some reporting on this. Instead of starting from scratch, I can ask Copilot to explain a visualization to me, and use that response as a starting point for my report.

Copilot has returned some key insights for us that would provide the most value given the visualization type. In this case, we see trends, outliers, consistencies, and more. 

Now, let’s take a look at the overall service level for this location. In February and March of 2023, we saw huge decreases from units shipped to units received, so I’m interested in the service level of that time period. I’m also interested in the most recent months, March and April of 2024.

For February and March of 2023, if we click on Intense Mocha, we see that National Shipping and Rocky Mountain Co Logistics transported that product, and they don’t have great service levels during this period. If we go back to viewing all products, it looks like Coffee Trucks Ltd has the highest carrier service level. 

Now looking at the most recent months, we see that Buffalo Transport, Regional Expediters, and Coffee Trucks Ltd are performing at a high service level. If we click on Intense Mocha again, we can confirm that National Shipping and Rocky Mountain Co Logistics still aren’t performing too well.

With these insights, I’m curious about some hypothetical scenarios, such as if we switched our primary carrier to Buffalo Transport, how many units can we expect to receive? To dive deeper into this, I’m going to ask Copilot to help me create a prediction data function.

With these predictions, we can gain some insights on how a different choice in carrier could improve our business outcomes. The last thing I want to do is get some final insights about how the carriers perform for the San Jose location specifically by creating some informative visualizations. Let’s look at the overall number of missing units across the different carriers and the amount of delays by carrier. The only adjustment I’ll make to Copilot’s output is limiting the data to look at the San Jose location.

Let’s also ask Copilot how to add a horizontal line to denote the average values for each visualization, then follow those instructions for both visualizations we just created.

Looks great. Through some digging with the help of Copilot, we now have a better understanding of the transportation issues that are happening, especially in regards to this location of interest. Now, I can take these findings back to my team, where we’ll be able to make an informed decision about what to do next.

Demo Recording

Check out the video below to watch the full demo!

Resources

• Spotfire Copilot Landing Page + FAQ
• Download Spotfire Copilot
• Spotfire Copilot: Interact with Spotfire in Human Language (history of our Copilot work)
• NLP and LLMs in Spotfire (overview of NLP and LLM tools for Spotfire)
• NLP glossary (understanding common terminology)
• Have questions? Need support? Contact us at datascience@spotfire.com.