The D3M textbook

NYU Stern · An open learning resource

Better questions. Stronger evidence. Better decisions.

Data Driven
Decision Making

From Business Questions to Visual Evidence, Algorithms, and AI Workflows

A practical textbook for managers and analysts. Learn to read the data, question the evidence, and turn analysis into action — with real cases and interactive studios along the way.

Vishal SinghNYU Stern School of Business

The question behind every chapter

What decision will this evidence improve?

From the classroom to practiceD3M
Connected parts
7
Chapters
18
Published articles
75

The curriculum

One book. Seven ways to think with data.

Follow the full arc, from business tables to AI agents. Or begin with the question on your desk. Each part brings together the concepts, worked cases, and practical tools you need to make the next decision.

  1. Part 0 · 1 chapter

    The Modern Data Operating System

    The map before the methods

    Explore this part
  2. Part I · 2 chapters

    Language of Data: Reading the Business in Rows and Columns

    Before any model, read the table

    Explore this part
  3. Part II · 2 chapters

    Visual Evidence: From Charts to Decisions

    Charts that answer, not decorate

    Explore this part
  4. Part IV · 4 chapters

    Language of Algorithms: Prediction, Segmentation, and Model Evaluation

    From explaining the past to predicting the next move

    Explore this part
  5. Part V · 4 chapters

    Unstructured Data, Embeddings, and Generative AI

    Turning prose and pixels into governed evidence

    Explore this part
  6. Part VI · 1 chapter

    D3M with AI Agents

    When the analyst is an agent

    Explore this part

Beyond the page

Get your hands on the evidence.

The ideas become useful when you put them to work. Explore the datasets behind the chapters, or try a hands-on teaching studio.

Explore interactive studios
Dataset portrait

Half a billion
Amazon reviews

33 product categories. 1996–2023. A guided look at the ratings, growth, seasonality, and biases hiding inside the corpus.

Open the portrait