Vibe Coding for Data Analysts
1h 2mIntermediate2025-08-14
Authors
Seth Berry
Associate Teaching Professor and MSBA Academic Co-Director at the University of Notre Dame
Course details
Get up to speed with vibe coding for analytics tasks in this course designed for data analysts. Instructor Seth Berry covers the essentials of setting up your environment for effective vibe coding, with the goal of creating an end-to-end application for hosting a machine learning model. From data import to model deployment, Seth highlights which aspects of the process are best left to humans and where LLM assistance can best be leveraged. By the end of this course, you’ll be equipped with the skills you need to create your own voice-assisted vibe coding project.
Learning objectives
Implement vibe coding strategies to perform common analytics tasks.
Understand and explain what phases of the analytics lifecycle are not appropriate for vibe coding and require human attention.
Create an end-to-end app for people to interact with your machine learning models.
Create a voice-to-code pipeline.
Learning objectives
Implement vibe coding strategies to perform common analytics tasks.
Understand and explain what phases of the analytics lifecycle are not appropriate for vibe coding and require human attention.
Create an end-to-end app for people to interact with your machine learning models.
Create a voice-to-code pipeline.
Skills covered
AI Development Tools and PlatformsProgramming FoundationsBuilding with AIData AnalysisData ScienceBusiness Analysis and StrategyBusiness Software and ToolsSoftware DevelopmentOne-Off
Concepts
Introduction
- Predict chart-topping hits with vibe coding
- Setting up your environment
Importing Data
- Loading data with pandas and PyArrow
- Loading data with Polars
- Reading data from HTML tables
- Generating SQL
- Merging data
Data Exploration and Wrangling
- Generating descriptive statistics
- Converting data types
- Examining correlations
- Aggregating data
- Reshaping data
- Creating derived variables
- A human s job - Knowing your data is ready
Data Pipelines
- Feature scaling and imputing data for scikit-learn
- Encoding categorical variables
- Partitioning data
- Creating data configurations for PyTorch Tabular
Model Creation and Testing
- Testing different baseline models
- Tuning models
- Creating a neural network with PyTorch Tabular
- A human s job - Evaluating performance
Visualizations and Dashboards
- Creating plots
- Creating explainer plots
- A human s job - Generating insight
- Building your model in a Streamlit application
- Deploying your model in a Streamlit application
Wrapping Up
- Security considerations
- Encouragement and caution