Product Builder, x-Shopify/Delphia YC18

Social posts give you fragments. Prompt libraries don't teach judgment. Structured practice is what builds skill.
After 12 hours (8 live), you will have shipped a working software product to your own GitHub repository from original research on a topic. Beyond vibe-coding is agentic engineering - a repeatable workflow for structuring work, research, tests, and deployment. You'll know what to delegate, how to check quality, and when to intervene.
Claude Code enables you to create specialists, and this course will teach you workflows to lead those specialists.
From the IDE, to Skills, MCPs, project management, and tokens needed for command line interface tools or APIs, this course sets you up.
By providing details, and constraints, agents aren't guessing, they are solving for the solution that fits the constraints.
Seeing the process through end-to-end gains new (human) skills
Production requires you to know take responsibility of live operations.

Ryan is an x-Shopify Product Leader, and prior deep learning founder
Senior leaders in software/IT needing to understand the new agentic AI capabilities in order to lead teams
Product Managers and Product Designers taking on more scope
Junior and intermediate software developers who want more structured examples that improve their craft.
Agentic AI workflows with Claude Code build on a foundation in software product research and development.
A minimum $100/month subscription is recommended to be able to execute tasks.
This course will require admin rights to install software on the computer

Live sessions
Learn directly from Ryan Smith in a real-time, interactive format.
The Agentic Product Builder Workflow
Set up your complete development environment and launch your first AI-powered research workflows. - Configure your toolkit: Claude Code, Linear, Vercel, VS Code - Establish security foundations with Docker and Varlock - Run AI-assisted user research, market research, and PRD generation - Create your first design brief using agentic workflows
Implementation with Specialist Agents
Learn to coordinate agents that write, test, and debug code. - Configure skills, MCPs, and subagents for specialized tasks - Implement test-driven development with AI assistance - Build and validate your golden data set - Debug systematically with agentic support
DevOps and Deployment
Master the Git workflows and deployment pipelines that get code into production. - Work with feature branches and pull requests - Set up preview, staging, and production environments - Deploy your working product to Vercel
Guardrails and Hardening
Prepare your product for real users with security, testing, and quality controls. - Red team your own implementation - Add guardrails for production reliability - Harden for customer-facing deployment
Maven Guarantee
This course is backed by the Maven Guarantee. Students are eligible for a full refund up until the halfway point of the course.
Live sessions
8 hrs
From research to DevOps in 8 hours, ideal for busy executives and leaders who need compressed agentic AI training. This session will focus on using Claude Code and a set of CLI tools to design, build, and ship software.
Fri, Jan 30
5:00 PMβ1:00 AM (UTC)
Projects
4 hrs
At least half the 8-hour session will be dedicated to operating Claude Code to write and deploy software.
Async content
1-2 hrs
Optional: The following Friday we will do a community session for this cohort to debrief next steps to help you gain momentum with agentic AI.
$899
USD