AI in Fabric - Copilot, AI Functions & Data Agents
Details
🎯 The Problem: Your boss says "just add AI to our reports." Your users want to "ask questions in natural language." But you're not a data scientist, and ChatGPT doesn't know your data. Where do you even start?
💡 This Evening: We'll show you the AI spectrum in Fabric - from traditional ML to conversational analytics - and which approach fits which use case. No PhD required.
⏱️ Save yourself: The frustration of picking the wrong AI tool. Copilot, AI Functions, and Data Agents solve very different problems.
What you'll learn:
🤖 ML Models - Train and deploy models in Fabric, and WHY MLflow is the backbone of Fabric ML (experiment tracking, model registry)
🧪 Experiments - Track runs, compare models, version artifacts, and WHY experiment tracking prevents ML chaos
✨ Copilot in Fabric - AI-assisted development across all workloads, and WHY Copilot context matters for quality (semantic model awareness)
🧠 AI Functions - Call LLMs from T-SQL and Spark, and WHY AI Functions democratize GenAI for SQL developers
💬 Data Agents - Conversational analytics with natural language, and WHY agents beat chatbots for data Q&A (they understand your schema)
AI Spectrum: Traditional ML (Train models) → Experiments (Track) → AI Functions (Call LLMs) → Data Agents (Conversational)
Who should attend: Data Scientists exploring Fabric, Analysts curious about AI capabilities, Developers building AI-powered apps
Agenda:
- 18:30 - Welcome & Networking
- 18:45 - ML Models & Experiments
- 19:10 - Copilot Across Fabric
- 19:25 - AI Functions Demo
- 19:40 - Data Agents - The Future of BI
- 19:55 - Q&A and Discussion
- 20:00 - Networking