Live Interactive Workshop

The Agentic Engineering Playbook

December 8 to 9, 2026 - 9:30am to 4:30pm Central Standard Time

Take a real project from plan to verified build. Scope work in Linear, design key screens, and direct the build. Then manage context in Obsidian, turn repeated tasks into skills, delegate to subagents, and match each task to the right model to keep costs down.

The Agentic Engineering Playbook
Meet Your Instructor

Software & Systems Architect

Steve is a Principle Engineer at 3Play Media. Previously, he was a Head of Engineering for Developer Tools at Temporal and a Frontend Architect at Twilio. In a previous life, Steve was a New York City public school teacher. He taught special education and web development in Manhattan, Brooklyn, and Queens. He currently lives in Denver, Colorado.

Plan, build, review, and verify one real project with AI agents

Take a real project from initial plan to verified build. You’ll scope the work into Linear issues, design key screens, and direct the build. As the project grows, you’ll manage context with Obsidian notes, turn repeated tasks into skills, and delegate independent work to subagents. Along the way, you’ll learn to control costs by matching the right model to each task, and you’ll leave with a reliable, repeatable system for shipping software.

What You Will Master

The Senior Engineer Toolkit
  • Turn a project idea into a markdown plan and scoped Linear issues with a reusable planning prompt and issue template.
  • Create a visual UI spec in Claude Design or Figma and attach it to the work before building.
  • Implement a well-scoped issue in a fresh agent session, adjust permissions and effort, and review the result in a browser.
  • Decide when to compact, fork, or start fresh from notes, and understand how those choices affect caching.
  • Turn repeated work into a skill that loads only when it is needed.
  • Use focused subagents for PR review and independent issues, with guardrails and a clear conclusion returned to the main thread.
  • Combine planning, implementation, review, and verification into a repeatable workflow with a model and cost for each step.

Prerequisites

  • Comfortable with Git, pull requests, and web application development.
  • Some hands-on experience with an AI coding assistant.

Who is this for?

For developers who already use an AI coding assistant and want a more deliberate way to plan, build, review, and verify real projects.

RSVP to Attend Online

Replay coming soon.

The Blueprint

Build a reliable, repeatable system.

Take a real project from plan to verified build. Scope work in Linear, design key screens, and direct the build. Then manage context in Obsidian, turn repeated tasks into skills, delegate to subagents, and match each task to the right model to keep costs down.

Warning: High intensity learning environment.

9:30AM
Introduction
10:00AM
Project scope & Linear issues
11:00AM
Iterate on design & visual specs
12:00PM
Lunch
1:00PM
Building with AI
2:30PM
Permissions, effort levels, and model costs
3:30PM
Obsidian, compaction, forks, and fresh sessions
9:30AM
Skills for repeatable tasks
11:00AM
Subagent PR reviews & triage
12:00PM
Lunch
1:00PM
Workflows for review & verification
2:30PM
Goals and loops for measurable tasks
3:30PM
Wrap up
Prefer to attend in-person?
Join us in downtown Minneapolis, MN. Limited seats available.