Explore natural-language analytics, curated data sources, trusted AI responses, security, governance, and how to design data experiences that help users ask better questions and get faster insights.
In Chat With Your Data in a Day, Greg Trzeciak shows how modern Microsoft data and AI tools can help users ask natural-language questions against trusted business data. This course is designed to help learners understand what “chat with your data” really means, how it differs from traditional reporting, and what needs to be in place for AI-powered answers to be useful, accurate, and grounded. The focus is on building a practical foundation for connecting users to data through conversational experiences.
Throughout the course, you’ll explore how curated data sources, semantic structure, security, and thoughtful design all contribute to a better chat-based analytics experience. Greg explains how to prepare data for natural-language interaction, evaluate the quality of responses, and understand where human review and governance still matter. By the end, you’ll have a clearer path for creating data experiences that let users ask better questions, get faster insights, and work more confidently with AI-assisted analytics.
Course Outline ( Free Preview)
Module 00 - Introduction
Greg introduces the course and explains what learners will build throughout Chat With Your Data in a Day. You’ll get a high-level overview of the tools, labs, and concepts that support AI-powered data interaction. This module sets expectations for how Copilot, data modeling, semantic preparation, and data agents come together.
Module 01 - What is Copilot
Learn what Copilot is and how it supports users in Microsoft data and productivity experiences. Greg explains the role of Copilot as an AI assistant and frames how it can help users ask questions, summarize information, and work more efficiently. This module provides the foundation for understanding the rest of the course.
Module 02 - Where is Copilot and What Does It Do
Explore where Copilot appears and what types of tasks it can support. Greg walks through the broader Copilot experience and explains how it connects to user workflows. You’ll learn how Copilot can assist with analysis, productivity, and data-driven interaction when the right information is available.
Module 03 - Initial Instructor Demo
Greg begins with an instructor-led demonstration to show what the final experience can look like. This module gives learners a practical preview of how data preparation, modeling, and AI features work together. The goal is to establish context before moving into the hands-on labs.
Module 04 - Lab 1 Start
Begin the first lab and prepare to explore core Copilot capabilities in the data experience. Greg introduces the lab goals, what learners will work on, and how this activity fits into the overall course. This module helps transition from overview concepts into hands-on practice.
Module 05 - Lab 1: Exploring Capabilities
Explore the capabilities available when working with Copilot and data. Greg guides learners through the features they will use to interact with data more naturally and efficiently. This module helps build familiarity with the experience before introducing more focused modeling and configuration tasks.
Module 06 - Lab 1: Filters and DAX
Learn how filters and DAX affect the way users interact with data and receive answers. Greg explains why calculations, measures, and filtering behavior matter when preparing data for AI-assisted analysis. This module helps learners understand that strong data experiences depend on reliable model logic.
Module 07 - Lab 2 Start
Start the second lab with a focus on improving the structure and usability of the data model. Greg introduces the next phase of preparation and explains how better naming, descriptions, categories, and related settings can improve the user experience. This module sets up the semantic clean-up work that follows.
Module 08 - Lab 2: Renaming
Learn how renaming fields and objects can make a data model easier for both users and AI tools to understand. Greg shows why clear, business-friendly names matter when people ask natural-language questions. This module helps learners improve clarity and reduce confusion in the data experience.
Module 09 - Lab 2: Descriptions
Add descriptions that give users and AI systems more context about the data. Greg explains how descriptions can clarify the meaning of fields, measures, and model elements. This module helps learners make the semantic model more understandable, searchable, and useful.
Module 10 - Lab 2: Data Category
Learn how data categories help identify the meaning and expected use of specific fields. Greg demonstrates how applying the right categories can improve interpretation and downstream experiences. This module helps learners make the model more structured and context-aware.
Module 11 - Lab 2: Summarization
Explore summarization settings and how they affect the way values are aggregated or interpreted. Greg explains why summarization behavior matters for accurate analysis and natural-language questioning. This module helps learners avoid misleading results by configuring fields appropriately.
Module 12 - Lab 2: Sort By
Learn how sort settings improve the way data appears in reports and analytical experiences. Greg shows how properly sorting fields can make outputs more logical and easier to interpret. This module helps learners refine the model so users see results in the expected order.
Module 13 - Lab 2: Synonyms and More
Enhance the model with synonyms and additional semantic improvements that support natural-language interaction. Greg explains how alternate terms can help users ask questions in their own words while still connecting to the correct data. This module helps make the model more flexible and user-friendly.
Module 14 - Lab 3 Start
Begin the third lab, focused on improving AI behavior through instructions and verified answers. Greg introduces the next set of configuration steps and explains how they help guide the user experience. This module sets up the shift from model preparation to AI response quality.
Module 15 - AI Instructions Part 1
Learn how AI instructions can guide the way Copilot or a data agent responds to user questions. Greg explains how instructions help define behavior, expectations, and response style. This module begins the process of shaping the AI experience so answers are more useful and aligned with the intended scenario.
Module 16 - AI Instructions Part 2
Continue refining AI instructions with additional guidance and examples. Greg builds on the previous module by showing how more specific instructions can improve consistency and reduce ambiguity. This module helps learners understand how thoughtful instruction design supports better user outcomes.
Module 17 - Verified Answers
Explore verified answers and how they can improve trust in AI-assisted data experiences. Greg explains how verified responses can help guide users toward approved, reliable answers for common or important questions. This module helps learners understand how to combine flexibility with governance.
Module 18 - Lab 4 Start
Start the fourth lab, focused on visualizations and standalone answer experiences. Greg introduces the goals for this section and explains how the next set of activities expands beyond model configuration. This module prepares learners to connect data interaction with clearer outputs and user-facing experiences.
Module 19 - Visualizations
Learn how visualizations support the Chat With Your Data experience by making insights easier to interpret. Greg explains how visuals can complement natural-language answers and help users better understand trends, comparisons, and results. This module reinforces the connection between AI responses and traditional data storytelling.
Module 20 - Standalone Verified Answers
Explore standalone verified answers and how they can provide approved responses outside a broader report or model interaction. Greg shows how this approach can help users access trusted information quickly. This module helps learners think about verified answers as reusable, governed knowledge assets.
Module 21 - Context Switching
Learn how context switching affects the user experience when moving between data, questions, answers, and supporting tools. Greg explains why maintaining context matters when users are exploring data conversationally. This module helps learners understand the importance of designing smoother, more connected workflows.
Module 22 - Productivity
Explore how Chat With Your Data experiences can improve productivity for users and teams. Greg explains how natural-language questions, guided answers, and better-prepared data can reduce time spent searching, interpreting, and manually building analysis. This module focuses on the practical business value of the overall approach.
Module 23 - Lab 5 Start
Begin the fifth lab, focused on creating and testing a data agent. Greg introduces the goal of building an agent-based experience and explains how the previous preparation work supports this next step. This module transitions the course from semantic preparation into agent creation.
Module 24 - Creating and Testing the Data Agent
Create and test a data agent that allows users to interact with curated data through natural language. Greg walks through the core build process and explains how to validate that the agent responds as expected. This module helps learners move from concept to a working agent experience.
Module 25 - Agent Instructions
Learn how instructions shape the behavior of a data agent. Greg explains how to guide the agent’s responses, define expectations, and keep the experience aligned with the intended use case. This module helps learners improve agent reliability and usefulness through clearer configuration.
Module 26 - Adding the Lakehouse Source
Connect a Lakehouse source so the agent can use curated Fabric data in its responses. Greg explains how adding a source expands what the agent can answer and why source selection matters. This module helps learners connect the agent to the data foundation needed for meaningful interaction.
Module 27 - Lakehouse as a Source
Explore how a Lakehouse functions as a data source for a chat-based analytics experience. Greg explains what the agent can use from the Lakehouse and how this supports natural-language exploration. This module helps learners understand the role of curated lakehouse data in building trustworthy data agents.
Module 28 - Publishing and Sharing the Agent
Learn how to publish and share the data agent so others can use it. Greg explains the steps involved in moving from a test experience to something accessible for users. This module helps learners think about deployment, audience, and usability beyond the initial build.
Module 29 - Testing the Agent in Standalone Experience
Test the data agent in a standalone experience to see how it performs outside the build environment. Greg demonstrates how to review responses, validate behavior, and identify areas for improvement. This module helps learners confirm that the agent is ready for real user interaction.
Module 30 - Limitations
Understand the limitations of Chat With Your Data and data agent experiences. Greg explains where AI may struggle, why source quality and model design still matter, and what users should be cautious about. This module helps learners use the technology responsibly and set realistic expectations.
Module 31 - Tenant Settings
Review the tenant settings that can affect Copilot, data agents, and related AI experiences. Greg explains why administrative configuration matters for enabling features, controlling access, and supporting governance. This module helps learners understand that successful implementation depends on both the build and the environment around it.
Module 32 - Conclusion
Greg wraps up the course by reviewing the major concepts covered throughout Chat With Your Data in a Day. You’ve explored Copilot capabilities, semantic model preparation, AI instructions, verified answers, visualizations, data agents, Lakehouse sources, publishing, limitations, and tenant settings. This final module helps learners connect the full workflow and identify next steps for applying it in their own organization.
Gregory Trzeciak holds a Master's Degree in Education from the University of Florida and brings over a decade of experience teaching at the high school, college, and professional development levels. Throughout his career, he has been recognized for his ability to make complex technical concepts accessible, engaging, and actionable for learners of all experience levels.
As a Senior Trainer at Pragmatic Works, Greg specializes in helping organizations adopt and get value from Microsoft Copilot and AI technologies. He works closely with executives, business users, and technical teams to transform AI from an interesting concept into a practical tool that saves time, improves decision-making, and increases productivity. In addition to his deep expertise in Microsoft Copilot, Greg is also an experienced instructor in Microsoft Fabric, Power BI, and the broader Microsoft data and analytics platform.
Known for his energetic teaching style and hands-on approach, Greg focuses on building confidence through real-world scenarios, personalized coaching, and practical application. His goal is to help every learner move beyond theory and successfully integrate AI and analytics into their daily work.
Outside of work, Greg enjoys fantasy football, walking his dog, and training for half-marathons.
