Advanced Techniques for building smart agents in Copilot Studio

This training is intended for:

• IT specialists involved in system integration, business process automation, conversational agent design, and AI solution implementation to increase organisational operational efficiency.

• DevOps engineers and solution architects who design, implement, and manage AI-based environments, using advanced tools such as Copilot Studio, Power Fx, connectors, and Dataverse integrations to ensure scalability and smooth operations.

• Business users familiar with Copilot Studio who want to expand their competencies in creating and customising agents that support business processes.

• Power Platform developers seeking to integrate Copilot Studio with Power Apps, Power Automate, and Dataverse to build comprehensive process automation solutions.

Business analysts planning to use AI agents for data analysis, process optimisation, and decision support within their organisation.

Comprehensive knowledge of Copilot Studio

You will learn all key platform features—from building simple agents to advanced scenarios with multiple agents and Dataverse integration. This will enable you to independently design solutions tailored to business needs.

• Practical implementation skills

We go beyond theory—through exercises and labs, you will learn to configure agents, create flows, and integrate them with other Microsoft services. This is knowledge you can immediately apply at work.

• Understanding responsible AI principles

You will discover how to apply security mechanisms, content moderation, and data protection in Copilot Studio. This ensures your solutions comply with best practices and regulations.

• Ability to personalise and optimise agents

You will learn to adjust the tone, style, and behaviour of agents, as well as select appropriate language models. This allows you to create more engaging and effective user interactions.

• Preparation for working with multimodal and autonomous agents

You will gain competencies in designing agents that handle various data types (text, image, voice) and operate autonomously, increasing process efficiency in your organisation

1. Building agents in Copilot Studio

   1.1 Designing conversational agents

   1.2 Integrating and utilising tools

   1.3 Enabling autonomous agent behaviours

2. Building multi-agent systems

   2.1 Scenario overview

   2.2 Transforming a single agent into a multi-agent system

3. Building autonomous agents

   3.1 Enabling autonomous agent behaviour with Event Triggers

   3.2 Implementing Adaptive Card notifications for real-time collaboration

4. Customising agent responses

   4.1 Large Language Models

   4.2 Agent models

   4.3 Response formatting

5. Applying responsible AI and content governance

   5.1 AI safety in Copilot Studio

   5.2 Content moderation

6. Implementing multimodal features

   6.1 Interacting with LLMs

   6.2 Multimodal prompts

7. Augmenting prompts with grounding

   7.1 Implementing the RAG pattern with Azure AI Document Intelligence and Azure AI Search

   7.2 Implementing Dataverse grounding for prompts

• Basic knowledge of Microsoft 365 environment and cloud services

• Ability to work with Power Platform applications (Power Apps, Power Automate)

• Familiarity with basic concepts of artificial intelligence and process automation

• Ability to work in a browser environment and use online tools

• Basic knowledge of data structures and databases (e.g., Dataverse)

• Familiarity with Copilot Studio, especially the ability to create topics and flows

•  manual in electronic form available on the platform: https://www.altkomakademia.pl/

• access to Altkom Akademia's student portal

Training method:

• Lecture (20%)
• Workshops (30%)
• Exercises (50%)

• Training: English
• Materials: English