Microsoft CoPilot/AI - Workshop

The first part of the workshop dealt with the use of Microsoft Copilot and its integration into various Microsoft products and services. The following topics were covered:

Introduction to Copilot: a description of Microsoft Copilot and how it can support organizations like Baumit by increasing productivity, fostering creativity while ensuring data privacy and security. Copilot in Microsoft 365: Copilot has been integrated into various Microsoft 365 products such as Teams, Outlook, Word, PowerPoint, Excel and others to make daily work more efficient. The functions range from generating texts to analyzing data.

Copilot extensions: Options were presented to extend Copilot with plugins and Microsoft Graph Connectors to integrate external data sources.

Data protection and security: A key aspect was ensuring data protection and governance in order to meet the requirements of data protection laws such as the GDPR.

GitHub Copilot and Azure: The integration of GitHub Copilot and Azure AI services, which are particularly important for developers, was also presented.

In summary, the presentation introduced Microsoft Copilot as a versatile tool that automates and optimizes various tasks in everyday work through the use of artificial intelligence.

The second part of the workshop was about an AI workshop covering various topics related to Artificial Intelligence (AI)

Here is a summary of the content:

Introduction to Artificial Intelligence (AI/KI): At the beginning, there is an introduction to the basic concepts of AI, including a brainstorming session to familiarize participants with how AI works and its principles.

Data as a foundation: The importance of data for the success of AI models is explained and that the quality of the data plays a crucial role. This is followed by examples of data sets and how neural networks work.

How does an AI calculate? A detailed section on how AI works, including computational methods and an explanation of ReLU and softmax functions.

Backpropagation and training methods: The training processes of AI are discussed, in particular backpropagation, a technique for minimizing errors in neural networks.

Case study: A practical example shows the development of an AI for recognizing handwritten numbers.

Different AI models and their application: Different AI models such as Convolutional Neural Networks (CNN), Generative Adversarial Networks (GAN) and Large Language Models (LLM) are presented, which are used in various application areas such as image processing and text generation.

Prompt Engineering: One section is dedicated to the art of “prompt engineering”, i.e. the creation of prompts in order to obtain high-quality results from AI systems.

Future of AI: An outlook is given on new trends and developments in AI, including agent systems and retrieval augmented generation (RAG).

Application examples: Various real-world application examples will be discussed to illustrate how AI can be used in different industries.