Today, we're tackling a topic you probably haven't seen before: how to forecast and make predictions in Microsoft Fabric using Generative AI and Large Language Models (LLMs).
While most videos focus on LLMs as assistants or copilots, I find that use case uninspiring and soon-to-be commoditized. Instead, I'm focused on logic intelligence—applying GenAI to core app logic, business, and data processing.
Timeline:
00:00 - Introduction
00:15 - Context and Common Use Cases of LLMs
00:39 - Assistance and Copilot Use Cases
01:25 - Introducing Logic Intelligence
02:11 - Examples of Logic Intelligence Use Cases
02:47 - Dataset Overview
03:59 - Prediction Use Case: Forecasting Tips
05:09 - Setting Up the Environment
06:07 - Generating Tables with LLM
07:07 - Loading and Preparing Data
07:57 - Creating Context with LLM for Predictions
09:42 - Summarizing Data for Predictions
10:06 - Analyzing Generated Tables
11:11 - Writing the Prediction Function
12:01 - Testing the Predictions
13:14 - Comparing Predictions with Mathematical Models
14:42 - Future of AI and Business Outcomes
15:00 - Conclusion and Call to Action
In this video, we'll cover:
- Setting up the environment and data for predictions.
- Using LangChain and Azure's OpenAI for data analysis.
- Predicting restaurant tips based on customer and waiter demographics.
- Comparing LLM-based predictions with traditional math-based methods.
We'll start with the setup, move to prompt creation, and show you how to leverage LLMs for accurate predictions.
If you're as tired as I am of hearing about assistants and copilots, and you're ready to see how GenAI can actually drive business outcomes, you're in the right place.
Remember to like, subscribe, and leave a comment if you have questions.
