{"id":3209,"date":"2024-12-09T08:00:00","date_gmt":"2024-12-09T16:00:00","guid":{"rendered":""},"modified":"2026-04-09T08:59:25","modified_gmt":"2026-04-09T15:59:25","slug":"delivering-your-supply-chain-copilot-getting-started-on-ideation","status":"publish","type":"ms-industry","link":"https:\/\/cm-edgetun.pages.dev\/en-us\/microsoft-cloud\/blog\/retail-and-consumer-goods\/2024\/12\/09\/delivering-your-supply-chain-copilot-getting-started-on-ideation\/","title":{"rendered":"Delivering your supply chain copilot: Getting started on ideation"},"content":{"rendered":"\n

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<\/a>” covered priority areas of ROI that a supply chain copilot can provide.\u202f<\/em><\/p>\n\n\n\n

Now that we\u2019ve explored priority areas of return on investment (ROI) that a supply chain copilot can provide in the context of our supply chain, let\u2019s 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\u2019s 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.\u202f<\/p>\n\n\n\n

\n
Explore Microsoft AI solutions<\/a><\/div>\n<\/div>\n\n\n\n

Technology requirements\u202f <\/h2>\n\n\n\n

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:\u202f <\/p>\n\n\n\n

    \n
  1. Data platform<\/strong>:
    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<\/a>. Other sources might include SQL databases such as Azure SQL<\/a>. 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<\/a>\u2014offering unified storage, experience, and governance\u2014Fabric was designed to be the data platform for the era of AI.<\/li>\n\n\n\n
  2. Analytics and AI<\/strong>:
    Some might say that \u2018this is where the magic happens\u2019, and when it comes to making sense of the data they would be correct. The analytics and AI layer is where predictions and detailed\u2014yet actionable\u2014insights are generated. As an example,     Azure AI Studio<\/a> empowers you to simplify the creation of AI-powered solutions, giving access to multiple AI techniques and approaches\u2014including generative AI with Azure Open AI<\/a>, but also incorporating tools such as AI vision<\/a> or AI Speech<\/a>.<\/li>\n\n\n\n
  3. Application layer<\/strong>:
    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<\/a> is something that can add significant value to a copilot. For example, when a supply chain copilot recommends actions to a supply chain planner\u2014for example requesting expedition of a shipment\u2014it would add so much more value if the copilot were able to execute the recommendation on behalf of the planner.\u202f This is where Azure Functions<\/a>, which are event driven, can play their part.\u202fTypically, this would require interaction with other execution systems through an API\u2014resulting in the need for careful API management <\/a>over the lifecycle of the different solutions.\u202f <\/li>\n<\/ol>\n\n\n\n
    \"Technology\n\t\t\t\n\t\t\t\t\n\t\t\t<\/svg>\n\t\t<\/button>
    Figure 1: Technology requirements for supply chain copilot.<\/figcaption><\/figure>\n\n\n\n

    Your own supply chain copilot: Design and delivery\u202f <\/h2>\n\n\n\n

    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.\u202f <\/p>\n\n\n\n

    Issues such as data sources, system integration, technology choices, and overall architecture can make the prospect of delivering your own copilots feel overwhelming.\u202f\u202f <\/p>\n\n\n\n

    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:\u202f <\/p>\n\n\n\n

    Rapidly ideating on value\u202f <\/h3>\n\n\n\n

    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:\u202f\u202f <\/p>\n\n\n\n