The follow-up

Step 1: Project setup

Open a terminal and start Claude Code:

cd ~/dev
claude

Paste this prompt:

Set up my project:
1. Create ~/dev/data-science/p2
2. Download the project materials from https://learnbydirectingai.dev/materials/datascience/p2/materials.zip and extract them into that folder
3. Read CLAUDE.md -- it's the project governance file

Claude will create the folder, download and extract the materials, and read through CLAUDE.md. That file has the full project context: the client, the deliverable, the tech stack, the ticket list, and the verification targets. It is the same kind of project governance file you used in P1.

Once Claude confirms it has read CLAUDE.md, you are set up.

Step 2: Read the follow-up email

Open materials/client-email.md. Wanjiku is back.

The chart you made in P1 -- no-show rates by visit type -- is on the wall behind reception. Grace printed it. The team understands the scale of the problem now.

But that chart looks backward. It shows patterns that already happened. Wanjiku is looking at tomorrow's schedule: 30 appointments, and she wants to know which ones are likely to be empty. If she knew that, Grace could call those clients with an extra reminder, or Wanjiku could double-book the risky slots.

That is the shift. P1 asked "what are the patterns?" P2 asks "who won't show up next?"

Step 3: The question shift

Open materials/project-plan.md. Read Section 1 (Problem framing).

The project plan frames a distinction you have not encountered before: description versus prediction. A descriptive analysis looks backward at what happened. A prediction looks forward at what will happen. The data is the same. The question is different. And that difference changes everything downstream -- how you prepare the data, how you build and evaluate the model, and what you deliver to Wanjiku.

This project follows a structured pipeline: problem framing, data preparation, model building, evaluation, communication. You will work through each stage in order. The project plan gives you the structure. The verification targets tell you what honest results look like.

Step 4: Reply to Wanjiku

Below the email, you will see reply options. Pick the one that fits -- something that confirms the prediction goal and mentions your plan to work through it systematically.

Wanjiku responds warmly. She mentions the chart is still on the wall, Grace loves it, and now she is excited about the next step. She confirms the data file is updated with three more months in the same format.

This is a returning client. You already know Wanjiku, her clinic, and her data. The work starts from where P1 left off.