Getting Started

  • Welcome & Overview
  • Logistics & Format
  • Team Participation
  • Prerequisites
  • Setup & Day 0 Checklist
  • AI Course Assistant (Aca)

Week 1 — Agentic Cowork Tools

  • Week 1 Overview
  • Week 1 Schedule
  • Week 1 Resources

Week 2 — Agentic Systems

  • Week 2 Overview
  • Week 2 Schedule
  • Week 2 Resources

Week 3 — Deep Learning

  • Week 3 Overview
  • Week 3 Schedule
  • Week 3 Resources

Week 4 — Foundation Models in Research

  • Week 4 Overview
  • Week 4 Schedule
  • Week 4 Resources
ESDS Advanced/Getting Started/AI Course Assistant (Aca)
OverviewESDS CourseAll Programs

Up next

Week 1 Overview

Overview and goals for Week 1

On this page

    © Garrett Graham and Denis Willett 2026

    AI Course Assistant (Aca)

    The course comes with its own AI Course Assistant named Aca — that's what the name stands for. It lives inside Claude, it knows this course, and it's there to help you learn — not to do the work for you. Many students arrive new to LLMs, so Aca is built to be warm, patient, and to meet you wherever you're starting from.

    What Aca is

    Aca is a Claude skill the course provisions to every student's Claude account. There's nothing to install — it's just there when your account is set up. A few things make it different from asking Claude cold:

    • It knows this course. Aca answers from the live materials on this site, so what it tells you matches what we teach.
    • It's tailored to you. Once you introduce yourself, it adapts its help to your role, experience level, goals, and the kind of data you work with.
    • It coaches, it doesn't hand over answers. Aca leans toward hints and questions first, so you build real understanding rather than copy-pasting a solution.

    How to use it

    You use Aca just by talking to it by name in Claude (Claude Desktop or Cowork). A few named commands get you started:

    • Aca init — a ~5-minute first-run interview that builds your personal profile. Run this once, near the start of the course, so everything after it is tailored to you.
    • Aca check-in — a short end-of-class reflection plus a few comprehension questions. There's an automatic daily check-in during Week 1 (weekdays, at the 11:50 AM ET class close, July 20–24, 2026), and you can run it manually any time.
    • Aca quiz me — extra practice questions on any day's concepts, whenever you want more reps.
    • Aca help — lists everything Aca can do.

    You can also just ask it course questions in plain language — "what time does class start?", "help me with the Day 3 exercise" — and it will pull the answer from this site.

    A worked example

    Here's a nice payoff: Aca is itself course material. On Day 3, the skills day, you'll open Aca's own skill file and read it as a worked example of how a well-built skill is designed. The helper you've been using all week becomes the thing you learn to build.

    What's recorded

    Transparency about what we capture

    The course passively captures Claude Cowork session telemetry — including the check-ins — via OpenTelemetry to an institute-hosted SigNoz instance, so instructors can see how the cohort is doing and improve the course. This capture is configured at the course/organization level; Aca itself sends nothing. Check-ins are reviewed at the cohort level to spot where the group is struggling — not as individual grades — and the quizzes are never scored against you. If you'd like to know exactly what's recorded, just ask Aca or an instructor.

    Reading the skill itself

    How Aca actually behaves lives in its own skill file, its SKILL.md — that's the authoritative description of how it works. Because the skill ships with your Claude account, you can open and read SKILL.md directly in Claude. Reading it is a Day-3-style exercise in its own right, and a great way to see a real skill up close. If you'd like the source another way, instructors can share it on request, and it's viewable in the Claude skill listing.