Practical AI Bootcamp

Empower your team with hands-on skills to unleash the power of Copilot in Excel

Real-World AI in 2025.

Is Your Team Ready?

Python in Excel is intended to empower everyone (and we mean everyone now that we also have Copilot in Excel with Python) to do data analysis in an easy, secure, and compliant way.”

-Keyur Patel, Microsoft Principal Program Manager

AI is Being Sold Like It’s Magic

Microsoft is promising that Copilot will magically transform your organization:

Everyone in the organization with Copilot in Excel can perform amazing data analyses.

Better yet, with Copilot in Excel, everyone can harness the power of Python to perform advanced analytics.

Best of all, with Copilot in Excel, nobody needs to understand analytics, and they don’t need to know Python!

What Microsoft Won’t Tell You

Copilot will make mistakes.
Sometimes, Copilot will even make up stuff out of thin air!

The technical term for this is "hallucinations."

Sam Altman, the CEO of the most famous AI company (OpenAI), stated that hallucinations are more of a feature of AI than a bug.

Think about this for a second. What Sam Altman is saying is that hallucinations are inherent in every AI. There’s nothing you can do about it.

Every AI will sometimes make stuff up at random.

How Your Team Owns AI

Because AI hallucinates, you still need a human in the loop to validate what Copilot produces.

If you don’t do this, you’re playing with 🔥.

That’s the bad news.

Here’s the good news.

With the right skills, your team can unleash the power of Copilot in Excel.

Practical AI Bootcamp

The Practical AI Bootcamp empowers your team with the skills they need to partner with Copilot in Excel.

The true power of Copilot in Excel comes from its “advanced analysis” mode. In this mode, Copilot can gain a deep understanding of your data and even automatically analyze it.

However, there’s a catch.

Copilot generates Python code to work its magic. This Python code typically uses battle-tested analytics techniques like:

  • Distributions

  • Correlations

  • K-means clustering

  • Random forest machine learning

As demonstrated in the video above, Copilot is always confident in the code it generates, even if it’s wrong.

The outcome?

Your team will have the knowledge and hands-on skills to truly partner with Copilot:

  • Guiding Copilot with well-crafted prompts.

  • Evaluating the quality of Copilot’s output.

  • Re-prompting Copilot when needed.

A library of Copilot in Excel prompts is included for your team to use and extend.

  • Preparation courses on Python and data analysis are included at no additional charge.

  • Targeted curriculum to unleash the power of Copilot in Excel fast.

  • Your team will build real-world skills via 12 hands-on labs.

  • Courses can be taught in-person or virtually. Choose what works best.

  • PDFs of all slides, including notes, are provided to your team.

  • Bundle additional courses to expand your team’s capabilities.

What professionals have to say

Bring the same high-quality training experiences of national conferences to your team.

Training delivered in person or live virtual - whichever works best.

“Very good course as an intro to machine learning. I feel that with what I learned today I can put these skills into practice at work.”

— David Green, EMWD

“Fantastic intro ML course that’s presented in an engaging way. The content was easy to understand and the labs were easy to follow along. I’ve left the course wanting to dive deeper into the topic.”

— Jessica Liu, O-I Glass

“MIND BLOWN…not by the difficulty of the class, but by how EASY Dave makes machine learning within the reach of aspiring Data Scientists.

Easily the highlight of this year’s conference for me. I feel empowered to bring this material back to the job, put it to use, and teach it to others.”

— Chet Phelps, Health Solutions

“Best training and instructor I’ve had. Organized, clear, good pace, helpful examples, and an engaging and fun instructor.”

— Alex Kurtz, Sourceability

“I am so glad to have started the conference in Dave’s class. He set a wonderful tone for what is yet to come. I hope my other courses measure up!”

— Christina Mitchell, Naphcare

“Great class! Engaging instructor. Wish I would have had more time this week to attend his other sessions.”

— Matthew Royalt, Southern Star Central Gas Pipeline

Practical AI Bootcamp Curriculum

3 days.

12 hands-on labs.

  • What is Machine Learning?

    Data Analyst, Teacher

    Why Decision Trees?

  • Course Datasets

    Exploratory Data Analysis (EDA)

    Hands-on Lab #1

  • Classification Tree Intuition

    Overfitting Intuition

    Gini Impurity

    Split Quality

    Splitting Categorical Data

    Splitting Numeric Data

    Classification Trees with Python

    Hands-On Lab #2

  • Under/Overfitting

    The Bias-Variance Tradeoff

    Supervising the Data

    Model Tuning

    Classification Tree Pruning

    Measuring Awesomeness

    Splitting the Dataset

    Modeling Tuning with Python

    Hands-On Lab #3

  • Feature Engineering Intuition

    Data Leakage

    Decision Boundaries

    Engineering Numeric Features

    Engineering Categorical Features

    Engineering Date-Time Features

    Missing Data

    Hands-On Lab #4

  • Regression Tree Fundamentals

    Numeric Feature MSE

    Categorical Feature MSE

    Feature Evaluation

    Tuning Regression Trees

    Imputation with Regression Trees

    Hands-On Lab #5

  • Bad, Tree! Bad!

    Ensembles

    Bagging

    Feature Randomization

    Random Forests with Python

    Hands-On Lab #6

  • Feature Importance

    Tuning Random Forests

    Model Testing

  • Additional Resources

    Wanna Kaggle?

    Copilot in Excel Prompts

The following is the 3-day curriculum. The curriculum can be expanded by bundling additional courses (see below).

Introduction to Machine Learning - Days 1 & 2

  • Course Expectations

    What is Cluster Analysis?

    Cluster Analysis Use Cases

    The Challenge of Clustering Data

  • The Iris Dataset

    The Hand-Written Digits Dataset

    The Heart Dataset

  • Hierarchical, Partitional, and Overlapping Clustering

    Prototype Clusters

    Density-Based Clusters

  • Introducing K-Means

    The K-Means Algorithm

    Euclidian Distance

    The Problem with Outliers

    Data Standardization

    K-Means Caveats

    Hands-On Lab #1

  • Evaluating Clusters

    Cluster Cohesion

    Evaluating Cohesion with the Elbow Method

    The Silhouette Coefficient

    Evaluating Clusters using the Silhouette Score

    Hands-on Lab #2

  • Introducing DBSCAN

    The DBSCAN Algorithm

    DBSCAN Caveats

  • Considerations for Optimizing DBSCAN

    Calculating min_samples

    Choosing the eps Value

    Introducing Nearest Neighbors

    Evaluating eps with the Elbow Method

    DBSCAN vs K-Means

    Hands-On Lab #3

  • Introducing Dimensionality Reduction

    Principal Component Analysis (PCA)

    PCA Concepts

    Hands-On Lab #4

  • The Problem with Categories

    Encoding Categorical Data

    Factor Analysis of Mixed Data (FAMD)

  • Supervised Learning Resources

    Cluster Analysis Resources

    Copilot in Excel Prompts

Cluster Analysis - Day 3

Course Add-Ons

Expand your team’s capabilities by bundling additional courses into your bootcamp.

All courses can be taught using Python in Excel.

  • Text Mining and Analysis

    This 1-day hands-on course is an introduction to the tools and techniques of transforming text data into a form suitable for analytics. Examples include clustering documents and sentiment analysis. Topics include tokenization, stemming, lemmatization, TF-IDF, and cosine similarity.

  • Logistic Regression Analysis

    This 1-day hands-on course teaches your team how to use logistic regression predictive models to analyze your data with the statsmodels library. Logistic regression is one of Copilot’s go-to techniques when the outcome of interest is in the form of a Yes/No question.

  • Linear Regression Analysis

    This 1-day hands-on course jumpstarts your team’s skills in using linear regression predictive models to analyze your data with the statsmodels library. Linear regression is one of Copilot’s go-to techniques when the outcome of interest is a numeric quantity (e.g., sales).

FAQs

  • Yes! The bootcamp can be delivered virtually or in-person with your team.

  • While three full days is ideal for learning, the Bootcamp can be delivered in a half-day format:

    Week 1 is four half days.

    Week 2 is two half days.

  • Yes! This bootcamp assumes your team has access to Copilot in Excel with Python.

  • While the courses do include mathematics, it is at a level accessible to a broad audience. For example, no knowledge of calculus or statistics is required.

  • All the courses use Python as the programming language. My Python in Excel Accelerator course is included and will teach your team what they need to know before the bootcamp.

  • My Visual Analysis with Python online course is included with the bootcamp. It will teach your team the foundational skills they need.