AI That Takes Action

Building Your First AI Assistants

Course Overview

Welcome to a course designed for everyone — especially those who think tech might not be for them! This course is your next step after learning the basics of AI chat. Instead of just asking AI questions, you’re going to build AI assistants that do things for you.

No coding. No technical background needed. Just a willingness to build, break, and try again.

What You’ll Learn

You’ll build real, useful AI assistants — things that save you time, handle repetitive tasks, and actually work in your daily life. By the end, you’ll understand what AI can realistically do for you, and you’ll have the skills to keep building on your own.

Course Format

5 sessions · ~2 hours each

“Your First Reusable AI Assistant”: Configure an assistant that works the same way every time
“Making Your Assistant Take Actions”: Connect AI to the tools you already use
“Memory — Making It Feel Like It Knows You”: Build an assistant that remembers your context
“When It Goes Wrong (and How to Fix It)”: Learn to diagnose and repair broken assistants
“Build Your Own”: Put it all together in a capstone project — and learn what it’s called

What Makes This Course Special

No Tech Background Needed: We explain everything in plain language as we go
Build First, Learn Why Second: You’ll have something working before we explain what it is
Real Problems Only: Every exercise is built around something useful in your actual life
Nothing Goes to Waste: Every session produces something you keep and actually use
Honest About Failure: We’ll break things on purpose — that’s where the real learning happens
Honest About AI: We talk openly about what AI does well and where it still falls short

Who This Course Is For

Anyone who has used ChatGPT or Claude and wants to go further
People who feel like they’re doing the same thing over and over and wish something could just handle it
Anyone who’s thought “tech isn’t really for me” — this course was made for you
Small business owners, community organizers, creatives, students — everyone
People who want practical skills, not theory

If you’ve already told an AI what to write, you’re one step away from telling it what to do.

Final Project

By the end of the course, each participant builds and keeps:

A reusable AI assistant configured for their own work or life
At least one automation that runs on its own
A personal prompt template they can reuse

Upon successful completion, you’ll earn:

Certificate of Completion with QR verification
Digital Badges for LinkedIn and professional profiles

Session 1: Your First Reusable AI Assistant

Most people use AI like a search engine — ask a question, get an answer, start over. This session introduces a different idea: an assistant you configure once that works the same way every time.

Key concept: Instructions (also called a system prompt) — standing guidance you write once that shapes every response.

What you build: A working assistant for a real task in your own life, configured with a role, a tone, and at least one constraint.

Milestones

✓ Have a working, configured assistant before the session ends

✓ Test it with real prompts from your own life

✓ Explain to a partner what role you gave it and why

Session 2: Making Your Assistant Take Actions

Your assistant from Session 1 can talk. This session gives it things it can actually do — send an email, log something to a spreadsheet, create a task. It stops just answering and starts acting.

Key concept: Trigger → action. Something happens, and your assistant does something in response — automatically.

What you build: A working automation connecting AI to at least one tool you already use, triggered by something that actually happens in your life or work.

Milestones

✓ Build an automation with at least one real trigger and one real action

✓ Run it with live input and confirm the output arrived where it should

✓ Describe the trigger → action loop in your own words

Session 3: Memory — Making It Feel Like It Knows You

Every time you open a new chat, the AI starts from scratch. This session changes that — you’ll build an assistant that remembers your context, your preferences, and your work.

Key concept: Two kinds of memory — standing instructions (always true about you) and saved context (specific things to refer back to).

What you build: A personal brief for your assistant and at least two pieces of saved context, tested against a real task that would have required re-explaining before.

Milestones

✓ Configure an assistant with a personal brief and at least 2 pieces of saved context

✓ Demonstrate it giving a better answer because of what it remembers

✓ Articulate one thing you chose not to share, and why

Session 4: When It Goes Wrong (and How to Fix It)

Every assistant eventually does something unexpected. This session reframes failure as useful information — and teaches you how to read it, diagnose it, and fix it.

Key concept: Two types of failure: bad instructions (vague, contradictory, missing edge cases) and missing guardrails (never told what not to do). Both are fixable.

What you build: A repaired and hardened version of a previous assistant, with at least one explicit guardrail added and tested.

Milestones

✓ Deliberately break an assistant and document what went wrong

✓ Successfully diagnose a partner’s broken instructions

✓ Add a working guardrail to at least one of your assistants

Session 5: Build Your Own (This Is Called an Agent)

Everything built over the past four sessions — an assistant with a role, that can take actions, that remembers things, with guardrails to keep it in check — has a name: an agent. This session puts it all together in a capstone build.

Key concept: An agent is an AI assistant that can take actions, use tools, and operate with some independence. You’ve already built several.

What you build: One complete agent of your choosing — from scratch, for your own work or life — presented to the group.

Capstone Requirements

Your agent must:

Have clear written instructions (role, goal, constraints)
Do at least one action beyond just answering (send, log, create, forward)
Remember at least one thing about you or your context
Have at least one guardrail

Milestones

✓ Build a complete agent that meets all four requirements

✓ Present it to the group with confidence

✓ Earn your Certificate of Completion and digital badges

Tools Used in This Course

AI Assistants

Claude Projects (claude.ai) — preferred for memory and instructions
ChatGPT Custom GPTs (chatgpt.com) — alternative for assistant configuration

Automation

Zapier (beginner-friendly, free tier available)
Make (more powerful, visual workflow builder — optional upgrade)

Memory & Storage

Claude Projects memory
ChatGPT Memory
Google Sheets (optional, for more structured data)

Digital Credentials & Certification Program

Certificate of Completion

Upon successfully completing all five sessions and the final capstone project, participants receive an official “AI Agents and Automation Certificate” that includes:

Verification Features: QR code linking to verified completion record
Date & Cohort: Official completion date and course cohort identifier
Digital Format: PDF certificate suitable for professional portfolios
Sharing Capabilities: Optimized for LinkedIn, personal websites, and resumes

Certification Requirements

✅ Attend & Participate: Complete all 5 sessions with active engagement

✅ Build & Submit: Submit your capstone agent and at least one automation from Sessions 1–4

✅ Present Work: Showcase your capstone in Session 5

✅ Provide Feedback: Engage constructively in peer review

Next Steps

After completing this course, you’ll be ready to:

Build AI assistants configured for your specific work and communication style
Create automations that connect the tools you use every day
Write clear instructions that produce reliable, consistent results
Diagnose and fix assistants when they behave unexpectedly
Evaluate new AI tools with confidence — knowing what questions to ask
Keep building, one useful assistant at a time