Delivering your supply chain copilot: Getting started on ideation

This blog is the second in a two-part series on delivering your supply chain copilot. Part one of the series “Delivering your supply chain copilot: Prioritizing areas of ROI” covered priority areas of ROI that a supply chain copilot can provide. 

Now that we’ve explored priority areas of return on investment (ROI) that a supply chain copilot can provide in the context of our supply chain, let’s delve into data considerations and how to get started on ideation for your supply chain copilot. The complex real-time decisions and agility required of today’s supply chains can seem daunting when thinking about transformation, but actually is well-suited to application of AI to selected use cases as you get started. 

Technology requirements  

Irrespective of the target areas for a supply chain copilot, the foundational technology capability requirements are the same. These can be thought of three distinct layers that ensure scalability and maximum extensibility into the future. The three layers are laid out below:  

1. Data platform:
   Data is the fuel that powers AI, consequently it must be the foundation of any application of AI such as a supply chain copilot. Data is likely to exist in multiple places, leveraging different storage approaches. Some could be in distributed databases such as Azure Cosmos DB. Other sources might include SQL databases such as Azure SQL. Irrespective of such sources, the need to unify across multiple locations is essential to make sense of the data. This need to govern, model, and consolidate multiple data sources creates a strong case for a single data platform like Microsoft Fabric—offering unified storage, experience, and governance—Fabric was designed to be the data platform for the era of AI.
2. Analytics and AI:
   Some might say that ‘this is where the magic happens’, and when it comes to making sense of the data they would be correct. The analytics and AI layer is where predictions and detailed—yet actionable—insights are generated. As an example, Azure AI Studio empowers you to simplify the creation of AI-powered solutions, giving access to multiple AI techniques and approaches—including generative AI with Azure Open AI, but also incorporating tools such as AI vision or AI Speech.
3. Application layer:
   While it is easy to think of the application layer as the layer with which users interact with your solution, this is only part of the story. Process orchestration and automation is something that can add significant value to a copilot. For example, when a supply chain copilot recommends actions to a supply chain planner—for example requesting expedition of a shipment—it would add so much more value if the copilot were able to execute the recommendation on behalf of the planner.  This is where Azure Functions, which are event driven, can play their part. Typically, this would require interaction with other execution systems through an API—resulting in the need for careful API management over the lifecycle of the different solutions.  

Your own supply chain copilot: Design and delivery  

Experienced supply chain and technology practitioners know very well that what may look amazing in a conceptual demonstration or video can be very difficult to deliver in the real world.  

Issues such as data sources, system integration, technology choices, and overall architecture can make the prospect of delivering your own copilots feel overwhelming.   

Microsoft teams have seen this and pioneered approaches to help you establish a way forward and deliver results quickly. This is broken down into three steps:  

Rapidly ideating on value  

This is primarily a workshop-driven approach to identify and validate specific challenges or opportunities where AI can deliver value for the customer. From a supply chain perspective, this would focus on the elements that you feel offer the largest potential benefit or cause the largest business pain. It is broken into three parts:   

• Inspiration: Here, a high-level education regarding the capabilities of AI, generative AI, and copilots are covered alongside inspirational examples of where customers are using this technology to deliver value.
• Opportunity generation: In this stage, opportunities within your supply chain organization are examined.
• Prioritization: This is the evaluation stage where opportunity areas are prioritized using a simple four-quadrant matrix with ‘Effort’ on the x-axis and ‘Impact’ on the y-axis. This provides a simple yet effective visualization approach to defining priority focus areas.  

The output of this is to be clear on the ”solution route,” or most appropriate combination of technologies to be applied, to deliver what is required. As an example, Microsoft Copilot Studio, Microsoft Azure AI Studio, and Microsoft Power Automate, when combined together, form a very powerful combination to support copilot delivery especially where there are requirements around information extraction, knowledge mining, and process orchestration.   

There is no “one-size-fits-all” so other approaches may be recommended. This may include leveraging partner offerings—Blue Yonder for example have mature capabilities in the form of its control tower solution which may be appropriate.   

Another key outcome of this stage is to understand which of four AI opportunities is being targeted: enrich employee experiences, reinvent customer engagement, reshape business process, or bend the curve on innovation. This aids identification of the key value drivers that you seek to influence, key performance indicators (KPIs) you seek to influence, and provides a segue into the next stage.  

Envisioning  

Again, this is a workshop driven approach—sometimes spanning several sessions—including all relevant customer stakeholders, to determine details of the solution to be delivered.   

This will include establishing a detailed view of architectural elements, interaction and integration points, value potential, and data requirements. A perspective on the business process impact will further enhance the detail behind the value case. This would leverage benchmarks from existing research and established copilots to create a view tailored to your business.  

An additional key output to complement the value case is a view of the investment required and the way in which the delivery can be structured to maximize return on investment and deliver using an Agile delivery approach.  

Proof of concept, minimum viable product, and beyond  

Following envisioning, there is a choice between moving to deliver a proof of concept (POC) or a minimum viable product (MVP).  

A POC demonstrates that an idea or use case is feasible through the delivery of a specific set of capabilities. It is used to illustrate and prove a concept and is usually self-contained in that it does not connect to live data or other systems. Consequently, it is of limited use beyond demonstrative purposes.  

By contrast, an MVP is deployed into production and integrated to existing systems so offers immediate value to the business and end-user while requiring limited effort to deliver. It can therefore become a foundation for further development and enhancement by adding capabilities based on prioritization using Agile development principles.

Integrate AI across your supply chain  

AI transformation, and specifically copilots, present a remarkable opportunity for you to innovate and compete with renewed vigor. By leveraging AI, businesses can enhance efficiency, mitigate risks, and uncover hidden opportunities.   

The integration of AI into supply chain management through a supply chain copilot, for instance, allows for real-time visibility, optimized data management, and seamless interoperation across multiple elements. This shift from reactive to proactive operations enables organizations to consistently deliver the right products at the right time, while balancing inventory, waste, and transportation costs. Moreover, the use of generative AI offers new possibilities for content and insight generation, further empowering supply chain practitioners. As technology continues to evolve, embracing AI transformation will be crucial for organizations to stay ahead in an increasingly complex and dynamic world.  

Learn more  

•  Microsoft Copilot Studio  
• Azure AI Studio   
• Microsoft Cloud for Retail website  
• Microsoft for Consumer Goods website  
• Supply chain AI for the new era of value realization  
• Unlock the full potential of your next-generation supply chain with Microsoft and Blue Yonder  
• Empowering responsible AI practices  
• Copilot in Microsoft Dynamics 365 Supply Chain Management   

The post Delivering your supply chain copilot: Getting started on ideation appeared first on Microsoft Industry Blogs.

Source: Microsoft Industry Blog