SQL for AI Projects: From Data Exploration to Impact

Course details

SQL plays a critical role in AI projects, allowing practitioners to explore data sources, validate business problems, and measure the impact of AI-powered experiments. In this course, Danny Ma—the CEO and Founder at Sydney Data Science—shows you how to identify AI opportunities and assess their feasibility. Explore SQL challenges that tackle natural language processing, customer recommendations, and website optimization in practical exercises that mirror common AI scenarios. Dive into the use of SQL for data exploration and experiment design and master the tools to efficiently transform data for AI models. Enhance your ability to convert business requirements into actionable AI tasks using SQL. Plus, learn about structured methodologies to measure AI model performance effectively and validate AI initiatives' impact on business outcomes. Whether you're a seasoned AI professional or new to the field, this course helps build impactful skills that contribute to successful AI project implementation.

Learning objectives
Discuss potential AI use cases and assess their feasibility from both a tech and commercial perspective.
Identify important data points from the user experience journey that could be used to AI use cases.
Explore the SQL database that contains the digital interactions and sales data.
Answer some key questions that can be used to validate the feasibility and overall opportunity for each AI use case.
Create multiple variations of a SQL data pipeline to generate training data for tabular machine learning and natural language processing AI models using window functions and array functions.
Assess the overall impact of AI experiments on user experience and financial outcomes.

Concepts

Introduction

Introduction

Data exploration for AI

Explore California case study
Business use cases for AI
Data exploration for AI
Measuring AI experiments

SQL for Natural Language Processing (Notebook 1)

Natural language processing 101
Define the problem
Data validation and cleaning
Getting Started with your Coding Environment
Cleaning HTML Text for NLP with SQL (DuckDB REGEXP REPLACE)
Data transformation
Transforming Text Data in SQL - Tokenization and more
Value measurement
A B Testing in SQL - Conversion Rates and more

SQL for Supervised Machine Learning (Notebook 2)

Supervised machine learning 101
Defining the current baseline
Generate ML inputs
Feature engineering
Preparing Data and Feature Engineering for Supervised ML
Designing ML experiments
Experiment Viability in SQL - Power, Uplift, and more

SQL for Website Optimization (Notebook 3)

Define the data problem
Data preparation
Value measurement
Measuring AI Performance - Metrics, Experiments & Insights

Conclusion

Next steps