Pawel Huryn · 10 min read · AI & Product
The Ultimate AI PM Learning Roadmap 2026
Read the original on productcompass.pm →
The article reframes tool comparisons (e.g., n8n vs Claude) as a false dichotomy and introduces a two-track AI PM roadmap: workspace agents (in your own workspace) and product agents (embedded in a product or process). It emphasizes learning foundational, tool-agnostic skills first, then applying them across both agent types, with Part 1 covering roles, concepts, prompt/context/intent engineering, and knowledge systems. The message: model changes every month, but core PM skills and system design endure.
Key Takeaways
- There are two agent categories: workspace agents (run in your environment) and product agents (embedded in a product or business process); you need both and use workspace agents to build product agents.
- Learn the foundations once, tool-agnostic, because monthly model changes don’t change the underlying skills you need.
- Core AI PM competencies include understanding the AI PM role, basic model capabilities/limitations, and designing around them (not just chasing new tools).
- Prompt, context, and intent engineering are essential skills that apply to both workspace and product agents, enabling reliable interactions and autonomous behavior.
- Knowledge systems connect how agents know things (memory, tools, retrieval), showing that the same concepts apply whether reading Claude Code or querying a production knowledge base.
Excerpt
Someone recently asked me: Did Claude replace n8n? Should you drop one and learn the other? I get why. For more than a year, when I wrote about building AI agents, I wrote about n8n. Then I started writing about Claude Code, Cowork, and Codex. But the answer is no. That is the whole point of this roadmap. You learn the foundations once. Then one question splits the rest: Does the agent run on your work, or inside your product? Some agents run in your workspace, on your own work: Claude Code, Cowork, Codex. Others are embedded inside a product…