AI Agent Launchpad · Small-group guided build

Build your own AI agent.Put it to work.

For your business. For your everyday work. Built around you.

A student building a personal workspace alongside a friendly AI agent.

Checking enrollment

—DAYS—HRS—MIN—SEC

Availability unavailable

Learn the pieces.Make something useful.

Customize your agent and build one useful workflow you can keep improving. A hands-on build, with guidance along the way.

8 live classes2-hour live class each week
10-person capacityA small-group guided build
October 22, 2026Starts at 6:30 PM Central

Your personal setup

Your machine. Your agent.A personal start.

One-on-one setup and a conversation about what you want to build.

A one-on-one AI agent setup and goal discussion.
Your personal setup

One-on-one.
Built around your goals.

Our first class introduces the course, what we will learn, and the tools in your tech stack. You will start by understanding your AI workspace and what it can become.

After that class, we will schedule a one-on-one session before class two. We will talk through your goals, what you would like your agent to work on or automate, and a useful first problem to solve.

Together, we will set up your Virtual Private Server with its own agent and the tools to get it running. We will verify access and make sure you know how to get back into your workspace.

Your starting stack

  • n8nYour workflow automation tool
  • DockerA home for your services in containers
  • HermesYour AI agent
  • Hermes WebUIYour browser interface for the agent
  • OpenAIThe models powering your agent’s reasoning

What you bring

A VPS you have purchased with Ubuntu 24.04 LTS and your own ChatGPT Plus subscription at $20/month. Keep both active for the duration of the course. You purchase and pay for them separately from the course fee.

ChatGPT Plus does not include separately billed OpenAI API usage. If your workflow needs the API, any API charges are separate.

Between classes 1 and 2 · Scheduled individually

Meet one-on-one for SSH provisioning and a goal conversation. Verify access, talk through your project, and arrive at class two with a working learning environment, not a done-for-you business system.

  • Your supported environment, ready
  • One practical task to focus on
  • A plan for your first workflow

Bring a VPS you have purchased with Ubuntu 24.04 LTS and a $20/month ChatGPT Plus subscription. You will need both for the duration of the course, and you pay for them separately from the course fee.

ChatGPT Plus does not include separately billed OpenAI API usage. If your workflow needs the API, those charges are separate.

Meet the instructor

Julian Fitzpatrick.Built with an agent. Taught by a person.

Julian Fitzpatrick, your AI Agent Launchpad instructor

I’m Julian, founder of BuildWithPXL. My background is in engineering production pipelines for television, film, commercials, and game cinematics. I help people, software, and complicated processes work together.

Today, I run BuildWithPXL with an AI agent embedded in my infrastructure. Together, we build custom tools around clients' needs, connect client portals and databases, create custom tools, and transform human workflows into agent backed workflows.

During the course, I’ll take you inside that larger agent build. You’ll see how the pieces connect, how I use them to run the business, and where human review keeps important decisions in my hands.

You’ll start by building a custom agent tailored to your needs. Then you’ll have a foundation you can keep growing in the direction you choose.

Want to talk through whether Launchpad fits what you want to build? Reach out, I’m happy to chat and answer questions before the course begins.

The journey

From idea toyour first workflow.

Start with a task you repeat. Connect only the information and tools you choose. Keep important decisions in your hands.

Your idea

Choose a task and the result you want.

Your agent

Follows your instructions and uses what you connect.

Your documents · Approved tools

Prepared work

A draft or result ready for you to check.

Your review

Check it. Request changes. Approve when ready.

Not approved? It stays here.

Approved action

Take the agreed next step and record the result.

Illustrative learning flow. Connections and approval controls are configured during your build.

Knowledge and tools

Bring in what your agent needs.Not everything.

A person organizing knowledge and approved tools around an AI agent.

Learn how to organize approved documents, check answers against the source, and connect a tool with care. You will test small before relying on a bigger workflow.

Your project can start with organized files. A database is not required.

Human review stays visible

Put it to work.Stay in control.

A person reviewing work prepared by an AI agent before taking action.

Build a repeatable workflow that prepares work and asks for your approval before important actions. A lead workflow, for example, can prepare a draft, but never auto-approve sending.

See what this course is, and is not

The eight-class journey

Learn and build in class.Make it yours each Week.

One project, built step by step. Eight live two-hour classes, weekly practice, and a working agent you know how to evolve at the end.

  1. Class 01

    Understand your AI workspace

    Learn what your tech stack is and what it can become.

    Weekly practiceWhat problem do you want your agent to solve?
  2. Class 02

    Make your agent yours

    Shape its role, working style, and personal context.

    Weekly practicePersonalize your agent and verify it in a fresh conversation.
  3. Class 03

    Define the task and design the sequence

    Choose one useful task and map inputs, rules, review points, and results.

    Weekly practiceCreate a workflow map and project brief.
  4. Class 04

    Add your knowledge and connect a tool

    Organize approved references and test one useful connection.

    Weekly practiceTest your information and connection separately.
  5. Class 05

    Build and schedule the workflow

    Prove the full manual path, then make it repeatable.

    Weekly practiceVerify a manual run before adding a schedule.
  6. Class 06

    Make your workflow reliable

    Find failures, test safe recovery, and learn when to pause.

    Weekly practiceDocument a known limitation and your recovery steps.
  7. Class 07

    Make it easy to use

    Organize results and make human review part of your routine.

    Weekly practiceTry a small authorized real task and improve one awkward step.
  8. Class 08

    Demonstrate, take ownership, and expand

    Show what you built, document it, and choose a practical next improvement.

    Weekly practiceFinish your ownership and maintenance checklist.
Your agent. Your workflow.
The knowledge to keep building.
Not eight disconnected tutorials, one useful system that grows with you.

Ownership

Leave with something you understand.And can improve.

A student confidently operating an organized AI workspace.

You own your VPS and model accounts. You leave with a customized agent, one useful connected workflow, supporting documentation, and practice operating it. Those accounts and their recurring costs are purchased separately.

Customized agentOne useful workflowSetup notes and operating guideNext-build plan

Fit and expectations

For people ready to build.Not just browse.

A good fit if you want to…

  • Learn by building around business or personal work.
  • Bring one repeated task and work on it over time.
  • Review results and make careful choices about access and actions.

This is not…

  • A finished custom platform or an autonomous employee.
  • A promise that every tool or account can be connected.
  • A replacement for purchasing and managing your own infrastructure accounts.

Join the next cohort

Build with a small group.Make it your own.

A calm mint and lavender workspace representing the next step after the course.

Focused live classes, weekly practice, personal setup, and time to make the work yours.

First 5 seats$997Save $253Next 5 seats$1,250

Ten seats, one small group. The first five paid seats are $997, then the next five are $1,250. Your seat is confirmed after successful payment.

VPS and model-account costs are purchased separately.

Back to the course overview